Patent Application
20050266920
The present invention relates to the processing of images in a virtually set-up game space (hereinbelow called “virtual space”) (hereinbelow called “virtual images”); in particular, it relates to image processing for a game whereby the movement of a virtual model that is set up in virtual space (for example called a “character”) on a screen can be represented more realistically. This image processing technique is suitable in particular for 3D video games machines.
Due to the development of computer graphics (CG) technologies, it has become possible to represent a virtual space (also called “virtual world”) that is set up in virtual fashion three- dimensionally and in real time. This is the technical field of video games machines: these have been developed in recent years and incorporate a central processing unit (CPU) capable of high speed computation and a video display processor (VDP) and enable such computer graphics technologies to be utilised at high speed and in economical fashion.
In such a video games machine, the content of game play changes continually depending on the actions of the user (also called the game player or player), so the objects that are displayed have to be moved freely at high-speed in virtual space. To this end, usually a model constituting an object to be displayed (for example a character) is constituted of a set of polygonal fragments called polygons of triangular or quadrilateral shape, and the movement of such models is displayed by simultaneously altering the spatial position of these polygons.
Also, when specified portions or faces of objects to be displayed such as the arms or legs of characters were to be moved simultaneously, a polygon data group consisting of an assembly of a plurality of polygons was taken as a unit and a spatial position was given to each polygon data group, so that the specified portions or faces could be moved simultaneously.
In recent years, so-called 3D (three-dimensional) games have attracted attention in the market, in which a character is made up of polygons and an image is represented on a monitor in which the movement of the character in virtual space is captured from a virtual viewpoint; simulations of fights between a plurality of warriors are particularly popular (for example, “Virtual Fighter” (trade mark) made by Sega Enterprises Limited). In such a fighting simulation game, a player rapidly operates a stick or pad or button attached to a controller to make a warrior represented on the screen perform actions in accordance with commands determined by operation of the stick etc. The movements of the warrior are called “motion” and data in order to implement such motion is acquired using the “motion capture” technique. This data is then processed, if need be, and is utilised as final motion data in a 3D video games machine. In such a video game, in order to raise the product value, it is desirable to represent the movement of the character more realistically. More specifically, this consists in for example adding more varieties of movement whilst approaching very closely to the actions of an actual warrior. However, since the anticipated movements of a character extend over a very wide range of possibilities, there are many problems that still need improvement in order to achieve such an objective. Of course, consideration has been given to for example compiling beforehand all the desired motion data and storing this in memory and obtaining characteristic expressions in order to get such motion, but the problem is that a large amount of data is required for this, which is impossible to process in real time.
A chief object of the present invention is therefore to provide an image processing technique for games whereby the amount of calculation and/or the amount of data required for image processing can be greatly reduced and whereby the movement of the character on the screen can be expressed more in real time and more realistically.
First specific aspects of a problem that was experienced in conventional video games machines from the point of view of this object are as follows.
A1. In a 3D video game, since the image on the two-dimensional screen from the virtual viewpoint is represented by performing a projection conversion, it is difficult to move the warrior in the depth direction of the screen (z direction of virtual space) i.e. in the direction of the player's gaze, so no consideration at all was given to enabling a warrior to walk around other warriors. To improve such movement around the characters is therefore a first specific object of the present invention.
A2. With a conventional 3D video games machine, there was the problem that, since the images were displayed from a virtual viewpoint, if a structure such as a wall was arranged in virtual space in a position such as to screen the warrior, display was effected with the character blocked out. Accordingly, a further specific object of the present invention is to effect display in which this situation that the character is blocked out by a structure is improved.
A3. In a conventional video games machine, the method was adopted of generating the motion of the character sequentially using for example a spline function, or the method of effecting reproduction of predetermined patterns in sequential frames. However, with the conventional video games machine, this motion was fixed, so it was not possible to correct the motion to match the movement of a character on the other side etc. Accordingly, a further specific object of the present invention is to enable such motion correction to be performed.
Further, as a derivation from the viewpoint of the above chief object, in addition to the first aspects described above, it is desired to perform screen display with improved dramatic effect such as character movement, in order to raise product value. Second aspects of the problems of conventional video games machines when this demand is taken into consideration may be described specifically as follows.
B1. “Motion blur” is known as a technique for improving the dramatic effect of a CG image. By means of such motion blur, a large number of rays are generated at a single pixel, and coloration is applied wherein these are averaged, thereby enabling a picture to be created that shows “out-of focus” or movement.
Furthermore, in recent years, in the field of CG video such as video games, in order to raise the dramatic effect further, in movement of a character, display is effected together with residual images such as are liable to be produced physiologically in human visual perception. For example, residual images may be attached to the track of a sword that is being waved by a warrior. Persons skilled in the art would therefore calculate polygons and to constitute residual images matching the motion of the warrior and display these residual images in the vicinity of the warrior.
However, the anticipated movements of characters are extremely diverse, so compiling polygons for a large number of modes of residual images matching all these cases and storing these in memory restricts the performance of a limited computer graphics device and furthermore calculation of residual-image polygons in conformity with the motion of the character puts a large load on the computer graphics device. Accordingly, in the formation of a CG image, yet a further specific object of the present invention is to make it possible to display residual images simultaneously with the actual image screen without large increase in calculation load (or more precisely, reducing the load), even though measures are taken to raise the dramatic effect.
B2. In a games device using a conventional image processing device, flying material such as sand or water splashes is displayed on the screen (for example “Sega Rally” (trade mark) manufactured by Sega Enterprises Limited). However, since such water splashes or sand scattering consisted merely in mapping texture on to polygons, it was not possible to reflect accurately the movement of the model (car etc.) by the flying material. Yet a further specific object of the present invention is therefore to reflect the movement of the model more accurately in the flying material.
B3. The quality of the conventional image processing device was insufficient in simulation of movement of a model falling in virtual space. Yet a further specific object of the present invention is therefore to raise the quality of simulation of movement of free-fall through virtual space.
B4. In a conventional image processing device, the proximity of a zone in virtual space and a moving model were determined and if this determination was positive, movement of the model was restricted such as not to go beyond the zone. Usually, as zones of this type, fixed-shape, for example quadrilateral or circular, type zones were set up. In this case, since the zone was of a typical shape, it was easy to calculate movement that would expel the character from the zone. However, when this zone was irregularly shaped, there was the problem that it was difficult to cope with this situation with the conventional computer graphics device. Accordingly, yet a further specific object of the present invention is to enable the calculation processing of the movement (motion) of a model after there has been a positive determination of collision between the zone and the model to be executed accurately and easily.
B5. Conventionally, in the field of image processing applied in games devices, when a series of periodic pictures (for example, a picture of repeating waves or skipping etc.) was to be reproduced, a series of such pictures was compiled by a manual operation to produce a texture series and a series of images in which the same action was repeated were represented by repeated sequential mapping of these on to a polygon.
Since the compilation of such texture series requires a lot of manual work, attempts have been made to utilise application software. A known example of such application software is called by the trade name: Alias/Wavefront (manufactured by Alias/Wavefront Inc. (110 Richmond Street, East Toronto, Ontario, Canada, M5c 1p1). When such software is employed, the target texture series can be obtained by supplying prescribed parameters to the software.
However, when prior art software of this type is employed, pictures in which the starting and ending of the texture series i.e. the texture pattern (mode) are not continuous for the player often result.
There was therefore room for improvement in the creation of repeating pictures in the field of conventional image processing. It is therefore yet a further specific object of the present invention to enable the compilation of a texture series of this type by image processing.
B6. In a prior art image processing device, there were scenes in which, in order to improve the dramatic effect of the reproduced image, slow reproduction was performed in which the reproduction speed of character action was slowed down. However, the benefits of such slow-reproduction performance were limited. Yet a further specific object of the present invention is therefore to provide means for image processing for games whereby the speed of reproduction can be varied in order to create a better dramatic effect.
A chief object of the present invention is that there are provided from various viewpoints techniques for the processing of images for games whereby movement of characters on a screen can be represented in more real-time fashion and more realistically by greatly reducing the amount of computation and/or amount of data required for the image processing. In the present invention, various constructions given below are adopted in order to achieve various specific objects derived from this main object.
A series of constructions constituting a first aspect of the present invention are as follows. Specifically, according to the present invention, there is provided an image processing device for games wherein a prescribed number of models are set up in virtual space and these models are controlled such that they move in prescribed directions in the virtual space and images of this virtual space from the virtual viewpoint are displayed by means for display, comprising means for image processing whereby virtual centripetal force is applied to the models. There is also provided a games device equipped with this image processing device for games. Furthermore, there is provided a recording medium on which is recorded information containing a series of procedures for implementing this means for image processing.
Furthermore, according to the present invention, an image processing device of this type for games comprises means for image processing whereby the amount of frictional force that is applied to these models is varied when the models are moving and when they are stationary.
Furthermore, according to the present invention, an image processing device of this type for games comprises means for image processing whereby a projection image of the models is displayed matching the surface shape of a stage on which the models are placed. Further, according to the invention, an image processing device of this type for games comprises means for image processing whereby a determination is performed of overlap of the field of view created by the virtual viewpoint with respect to the model which is the subject of display and another model which is not set as the subject of display, and, if the result of this determination is affirmative, the other model is not displayed.
Further according to the present invention, there is provided an image processing device of this type for games comprising means for image processing whereby, with respect to the end-point of a predetermined movement track of the model, a target point is set up different from this end point and the movement track is interpolated such that the end point coincides with this target point.
Further according to the present invention, there is provided an image processing device of this type for games comprising means for image processing whereby a difference of level at which the model is positioned in the virtual space is found and the action which is applied to this model is interpolated in accordance with this level difference.
By means of such constructions, the movement of the model such as a character that is arranged in virtual space can be displayed more realistically. Specifically, images whereby one model moves around another model can be easily generated. Furthermore, the condition of the stage on which the model is placed can be reflected in the movement of the model. Furthermore, unevenness of the stage can be reflected accurately and in a simple manner in the projection image of the model. Furthermore, even if a structural object is a wall arranged in virtual space at a position which would screen the model, display can be effected without the model being screened. Also, versatile movement of the model can be represented in a reliable manner since it is possible to correct motion of the model by matching it to movement etc. of the other model. Furthermore, more realistic images can be generated since it is possible to take into account level differences between models.
Also, a series of constructions that constitute a further aspect of the present invention is disclosed below. As one specific construction thereof, according to the present invention there is provided an image processing device for games wherein a prescribed number of models are set up in virtual space and these models are controlled such that they move in prescribed directions in the virtual space and images of this virtual space from a viewpoint of prescribed position are displayed by means for display; comprising means for residual image representation processing for representing the track of movement of a model as residual images; this means for processing comprising means for storage whereby motion data of the model before the current motion are stored without modification and means for display control whereby this stored data is displayed together with current motion data of the model. This means for display control may comprise means for dramatic effects for effecting display with addition of dramatic effects to the data of the means for storage. This means for dramatic effects may perform semi-transparent image processing.
As a further specific construction, according to the present invention, there is provided an image processing device for games wherein a prescribed number of models are set up in virtual space and these models are controlled such that they move in prescribed directions in the virtual space and images of this virtual space from a viewpoint at a prescribed position are displayed by means for display; comprising means for dramatic effects that apply dramatic effects to the movement characteristic of the model and means for controlling flying material that cause flying material to be present in the virtual space and that control the series of movements of this flying material in accordance with the results of the calculation by the means for calculation. As a further specific construction, according to the present invention, there is provided an image processing device for games wherein free-fall movement of a model set in virtual space is simulated and comprising first means for simulating overall movement of the model and second means for conferring a circulating movement on the model. Further, as another specific construction, according to the present invention, there is provided an image processing device for games wherein a prescribed number of models are set up in virtual space and these models are controlled such that they move in prescribed directions in the virtual space and images of this virtual space from a viewpoint at a prescribed position are displayed by means for display; comprising means for setting up a zone that set up an irregularly shaped zone in the virtual space; means for setting up a vector consisting of means for setting up a prescribed vector between this zone and this model and that, if this model tries to move beyond this zone, sets up this vector in a direction such that the model does not go beyond the zone; and means for controlling model movement that control the movement of the model in accordance with this vector.
Furthermore, as a further specific construction, according to the present invention, in a device in which texture is formed that varies in periodic fashion on application of a prescribed parameter, there are provided first means for creating a first texture series and means for creating a second texture series and a texture series is formed wherein parameters are applied to the respective means such that the first or last texture of the first texture series and the first or last texture of the second texture series are constituted in continuous mode and the degree of transparency of at least one of the first texture series and second texture series is gradually increased or decreased with the object that superposition is effected in order from the first textures of both texture series.
Also, according to the present invention, in a method of forming texture series wherein periodically changing textures are formed, there are provided a first step of creating a first texture series which changes in a prescribed direction and a step of creating a second texture series that likewise changes in a prescribed direction and wherein the last or first texture of the first texture series and the first or last texture of the second texture series are made to be constituted in continuous mode and the degree of transparency of at least one of the first texture series and second texture series is gradually increased or gradually decreased with the object that superposition is effected in order from the first textures of both texture series.
Further, as a further specific construction, according to the present invention, there is provided an image processing device for games wherein a prescribed number of models are set up in virtual space and these models are controlled such that they move in prescribed directions in the virtual space and images of this virtual space from a viewpoint in prescribed position are displayed by means for display; comprising first means that determines whether or not a prescribed condition is established between two models; second means for reflecting the establishment of the condition in the movement of the respective models and third means for, if this condition is established, changing the reproduction speed of movement of one model with respect to the other model.
By means of such constructions, the movement of models in virtual space can be represented more realistically and games images that are rich in dramatic effects can be generated. In more detail, with one specific construction, there can be provided an image processing device for games in which large calculating load is not applied even though measures are taken to improve the dramatic effects in forming game images. Also, with another specific construction, there can be provided an image processing device for games whereby, even though such load is small, residual images can be displayed simultaneously with the real image screen. Furthermore, according to a further specific construction, there can be provided an image processing device for games wherein movement of a model can be accurately reflected in flying material. Further, with a further specific construction, there can be provided an image processing device for games whereby simulation of higher quality of free-fall movement in virtual space can be achieved. Further, according to a further specific construction, there can be provided an image processing device for games whereby calculation processing of movement (motion) of a model after an affirmative determination of collision between a zone and a model can be executed easily and reliably even for the case of an irregularly shaped zone. Further, according to a further specific construction, a series of texture series can be created whereby continuous images can be represented without a feeling of incompatibility even though display is repeated and image processing is employed. Yet further, with a further specific construction, there can be provided an image processing device for games in which the speed of reproduction can be varied to create a better dramatic effect.
Preferred embodiments of the present invention are described below with reference to the drawings, in respect of a video games machine constituting an image processing device.
A first embodiment of a video games machine is described with reference to
Description of the Hardware
This video games machine comprises: a CPU block 10 that performs control of the device as a whole, a video block 11 that performs display control of the game screen, a sound block 12 that generates effects sounds etc., and a subsystem 13 that performs reading of a CD-ROM.
CPU block 10 comprises an SCU (System Control Unit) 100, main CPU 101, RAM 102, ROM 103, sub CPU 104, and CPU bus 105 etc. This block is the heart of the image processing device of the present invention. Main CPU 101 incorporates a DSP (Digital Signal Processor), which executes computer programs at high speed. RAM 102 stores polygon data of various types that are sent to it from subsystem 13 that reads the CD-ROM, and is also employed as a work area of the main CPU 101.
ROM 103 stores an initial program for initial processing in the initial condition of the device. SCU 100 governs the transfer of data that is performed through bus 105, 106 and 107. Also, SCU 100 incorporates a DMA controller that sends the image data required during running of a game to VRAM in video block 11.
Control pad 2b functions as means for information input by the user and is equipped with various buttons needed for operation. Sub CPU 104 is called an SMPC (System Manager & Peripheral Controller) and has the function of collecting peripheral data from controller 2b in accordance with requests from main CPU 101.
Main CPU 101 performs processing such as movement of an image displayed on a display in accordance with peripheral data sent to it from sub CPU 104. Sub CPU 104 identifies the peripheral equipment that is connected to connector 2a (terminal of the main unit) and collects peripheral data in accordance with the communication system corresponding to the type of peripheral equipment that has thus been identified.
Video block 11 acts as means for figure generation and comprises a VDP (Video Display Processor) 120 for generating an image displayed using polygons and a VDP 130 that performs image synthesis, shaded surface processing, and clipping for the background image. It is connected to VDP 121 and frame buffers 122, 123.
When an image to be displayed on the display is generated in virtual space, the polygon data necessary for display are sent from CPU 101 through SCU 100 to VDP 120 and are written to VRAM 121. The polygon data written in VRAM 121 are stored in image-drawing frame buffer 121 or 123 as image-drawing data including colour information of 16 bits or 8 bits per pixel. The image-drawing data that are stored are sent to VDP 130. Main CPU 101 supplies control information to control image drawing through SCU 100 to VDP 130. VDP 130 controls the image-drawing data in accordance with this control information.
VDP 130 is connected to VRAM 131 and comprises a scroll function whereby the entire display screen is shifted vertically or horizontally or is rotated and a priority function (Z sort or Z buffer) that determines the order of display of polygons. VDP 130 outputs image-drawing data through memory 132 to encoder 160. Image-drawing data output to encoder 160 is converted to video-signal format and is then subjected to D/A conversion and displayed on monitor device 5. An image is displayed on monitor device 5 based on this video signal.
Sound block 12 comprises a DSP 140 that performs audio synthesis by the PCM system or FM system and a CPU 141 that controls DSP 140. The audio data generated by DSP 140 is output to two speakers 5a or 5b by D/A converter 170 after conversion to a two-channel signal. Subsystem 13 includes a CD-ROM drive etc. and is equipped with a function of reading application software supplied by a recording medium such as a CD-ROM and a generating video etc.
The processing performed by an image generating device as described above will now be described.
Circular Movement Processing:
In general terms, the effect of this processing is as follows. When a direction key 26b (also see
The details of this processing will now be described with reference to the flow chart of
Next, an angle in this direction is applied to the player's warrior 30 and virtual centripetal force towards the other side's warrior is applied to this warrior (step 404). Considered physically, this centripetal force is equivalent to an attractive force acting between the two objects.
If the x co-ordinate and z co-ordinate of the other side's warrior are respectively taken as (exp, ezp), and the x co-ordinate and z co-ordinate on the other side's warrior are respectively taken as (mxp, mzp), the direction of the other side's warrior 32, as described above, is calculated from: arctan (exp−mxp, ezp−mzp). For convenience, the y-direction co-ordinate i.e. the height direction of the warriors on both sides is neglected in this calculation.
As a result, circular motion in virtual space is produced (step 406) in accordance with the speed in the horizontal direction produced by the transverse direction key and the centripetal force. By means of this circular motion, the player's warrior 30 is made to move around the other player's warrior 32 whilst facing the other player's warrior.
Consequently, with the circular movement processing described, a single character is given a virtual centripetal force, so the player, simply by operating direction key 26b of
Virtual Friction Processing:
The effect of this processing is that the movement of the character is made more varied by altering the slippage in the movement of the character depending on the slope of the ground surface and the frictional force acting on the character in accordance with the nature of the ground surface, and applying dynamic friction when the character is moving and static friction when the character is not moving.
First of all,
Amount of slippage(=dv/dt)=ivg
where
As shown in
Next, in step 802, a determination is made as to whether the character (warrior) is moving or not. This is done for example from the point of view of the player by determining whether or not the direction key for controlling movement of the warrior is being operated (determination 1).
If the character is moving, dynamic friction is applied; if the character is not moving, static friction is applied (steps 804, 806). The friction (=dv/dt) itself is calculated as follows.
M·(dv/dt)=μ·S·M(g·cos θ+(dv′/dt)·sin θ)
dv/dt=,T·g.cos θ+T·(dv′/dt)·sin θ)
where M is the weight
The dynamic friction and static friction are determined by applying prescribed processing to the friction (or coefficient of friction) obtained by this expression. The static friction has a larger value than the dynamic friction. This friction changes depending on the angle of the ground surface, the friction being smaller as the slope of the ground surface is increased.
Next the attributes of the ground surface are determined. That is, it is determined whether the surface on which the character is grounded is for example a water location or a sandy location. These attributes are applied virtually beforehand to each ground surface. For example, the amount of friction calculated in the case of a sandy location is taken to be doubled and the amount of friction in the case of a water location is taken to be tripled. This is then reflected (step 810) so as to subtract this frictional force from the amount of slippage that was previously calculated. Thanks to this processing, static friction or dynamic friction is constantly applied to the character, so once a character has stopped it will not readily move off again, but once it has started moving it becomes easier to move. Also, the movement characteristics of the character can be altered depending on the attributes of the ground. By this means, movement of a character can be represented very realistically.
Display of shadow on to a ground surface having a level difference: As shown in
E′(x′,y′,z′)=Ms.E(x,y,z)
where E′(x′,y′,z′) are the co-ordinates of the shadow.
This Ms is given as follows.
Ms=Mu.(−Tyg).Pn.Tyn.(−Tsn).Rn.Tsn
Mu is a conversion matrix for converting the co-ordinates of the character to the world co-ordinate system of
Tyg is a parallel movement matrix for producing parallel movement from the ground surface on which the character is standing up to the base line.
Pn is a conversion matrix for slantwise projection of the character on to the base line.
Tyn is a matrix for effecting parallel movement of the slantwise projected shadow to the ground-contacting point of each object (i.e. the ground surface where each object is located).
Tsn is a parallel movement matrix for effecting parallel movement of each object with respect to the origin of the world co-ordinate system.
Rn is a matrix for rotating shadow polygons in accordance with the slope of the ground.
This processing is carried out by the following steps.
Consequently, by this processing, if there is irregularity in ground surface 90, the projected image of a character can be displayed in a simple manner on ground surface 90 incorporating such irregularity in a manner matching this irregularity.
Intersection Determination Processing:
In
r{right arrow over ( )}=R{right arrow over ( )}−L{right arrow over ( )}
t{right arrow over ( )}=T{right arrow over ( )}−L{right arrow over ( )}
p{right arrow over ( )}=c{right arrow over ( )}+L{right arrow over ( )}−T{right arrow over ( )}
cx=rx.|c{right arrow over ( )}|/|r{right arrow over ( )}|, cy=ry.|c{right arrow over ( )}|/|c{right arrow over ( )}| [Mathematical expression 1]
where:
|c{right arrow over ( )}|=|t{right arrow over ( )}|cos θ [Mathematical expression 2]
From the definition of the inner product,
|r{right arrow over ( )}|=|t{right arrow over ( )}| cos θ=rx tx+rz tz
|c{right arrow over ( )}|/|r{right arrow over ( )}|=(rx tx+rz tz/|r{right arrow over ( )}|2=(rx tx+rz tz)/(rx2+rz2) [Mathematical expression 3]
where
0≦|c{right arrow over ( )}|/|r{right arrow over ( )}|≦1 [Mathematical expression 4]
If the centre T of the approximation circle is within the positive region of straight line TR (if T is outside the inside of the triangle), if
|p{right arrow over ( )}|−d>0 [Mathematical expression 5]
the circle is independent with respect to the region within the triangle i.e. it does not overly this region; if
|c{right arrow over ( )}|/|r{right arrow over ( )}|>0 or
1<|c{right arrow over ( )}|/|r{right arrow over ( )} [Mathematical expression 6]
if point T is in the positive region of straight line LR, the circle is independent with respect to the region.
In step 110, if the approximation circle does not overly the internal region of the triangle, processing is performed (step 111) such that this structural object is not displayed; if it does not overly this, processing advances to the next step (step 112). The effect of the next processing is to perform a determination as to whether there is overlapping in respect of a structural object such as a wall where this structural object has length in a fixed direction. This processing is performed as follows as shown in
F1(P)=a1 Px+b1 Pz+c1 [Mathematical expression 7]
If the function for determining the positional relationship of straight line ST and an arbitrary point P is taken to be F2 (P),
F2(P)=a2 Px+b2 Pz+c2 [Mathematical expression 8]
If F1 (S)·F1(S)<0 and F2 (S)·F2(S)<0, it is determined that straight line TS intersects straight line LR.
If this intersection is denied, processing shifts to step 114 and the determination is made to whether there is overlap with the approximation circle. The reason for this is as follows. If the straight line TS described above enters fully within the region of the triangle, the result of the overlap decision is negative and the wall-shaped structural object would be displayed.
Accordingly, a small circle 150 corresponding to the thickness of the wall as shown in
By such processing, it can be arranged that structural objects such as walls, fences, or screens in the field-of-view region from the virtual camera to the character to be displayed are not displayed. However, if, as shown in
Character Motion Processing Control:
The effect of this processing is to form interpolation such that the final position of a single motion is matched to a target point, when it is desired to effect display in series from a single motion without linkage of other motions.
This action of the arm part is displayed on the monitor by calculation by means of a well-known spline function (in particular, a third-order spline function). If now at this point it is assumed that this arm part 160A is moving towards the breast part of the other warrior, and, at a certain time point, the other warrior moves in the y direction (for example, suddenly stoops), from the point of view of the player, it is very difficult to suddenly change the track of the moving arm to downwards as in
Accordingly, a link region is provided in the vicinity of the termination of the spline track alteration process of the motion that is currently being executed, and interpolation such as applying a new spline function is performed in this link region such that the terminal point of the initial spline track becomes a target point in another position.
This will now be described with reference to the drawings. As shown in
Let us now assume that fist 180A is moving from the starting point of spline function track 180 towards its end point. Let us assume that, in this process, due for example to a change in the behaviour of the other warrior, the need arises to alter the initial end point to a target point.
Accordingly, at the time point where the position 180A of a fist is in the link region, CPU block 10 corrects the function in this region such that it terminates at the target point. This interpolation is performed for example as follows. Of course, this could also be achieved by suitably altering the degree and/or coefficient of the spline function in the link region. As a result, the motion shown in
The link region (link time) is appropriately set up by means of a group of a plurality of frames comprising an end frame. Also, interpolated motion of the entire arm is effected by applying similar processing for the objects other than the fist, namely the upper arm part and lower arm part. The interpolated co-ordinates in the link region are set by means of the following characteristic expression:
Current co-ordinate+(current co-ordinate−target co-ordinate)·(Current time/link time)
In this process, the link frames constituting the link time represent 0.1 of the maximum number of frames of the motion (number of frames from the start frame of
The following benefits are achieved by this processing. When, as in a video games device, the controller is operated in order to rapidly operate a character such as a warrior appearing on the game screen, varied movements of the character are continuously reproduced. At a time point when a series of motions is reproduced, even though [the player] tries to alter the motion to match rapid actions of the other character, so long as the motion itself is fixed, this is difficult. However, if motion is interpolated as described above, the motion matching the movements of the other character for example can be reproduced in a varied manner, making the motion more realistic. Usually, such interpolation is extremely difficult for a user to perform and conventional video games devices are particularly ill-constructed to performing such interpolation.
It should be noted that, while this processing has been described taking as an example the case where the motion was calculated using a spline function, this processing could be applied to pattern change systems in which a predetermined pattern is sequentially reproduced. In such cases, it may be arranged to reproduce a corrected pattern in the link region.
Character Level Difference Processing:
The effect of this processing is to display a desired image in which this level difference is corrected when there is a mutual level difference between a character (warrior) and the ground. For example, as shown in
In this processing, as shown in
If the result of this determination is negative, next, processing returns to the main routine. On the other hand, if the result of this determination is positive, further, in step 2102, a collision determination is performed between the character and ground i.e. to establish whether the other warrior is standing on the ground; if the result of this is positive, character level difference processing is deemed to be necessary, and processing shifts to the next step. If this determination leads to a negative result, the routine in the Figure is terminated.
Next, in step 2104, the distance from the ground on which the character controlled by the player is standing to the aforementioned character is calculated. The same calculation is performed for the other character also. Next, in step 2106, the value of this result is compared with a prescribed value, and, if the range of the prescribed value is exceeded, processing is performed to keep the difference relating to this distance within this range. Specifically, if we let the height of a warrior operated by one player be (my) and the height of a warrior operated by another player or operated automatically by a prescribed program by the image processing device itself be (ey), the difference (diff1) between these two is (my−ey). An evaluation is then performed as to whether for example diff1 is within the range −0.15≦diff1≦0.15, and, if diff1 exceeds this range, a diff1 smaller than this range is taken as being −0.15 and a diff1 greater than this range is taken as being 0.15, while a diff1 within this range is directly determined as (diff2) without modification. The reasons for carrying out this correction on diff1 will be described later.
In the next step 2110, diff2 is divided by the time (number of frames) for which attacking force is generated and the result is taken as “result 1”; the result obtained by dividing diff2 by the time from generation of the attacking force until hardening is dissolved is taken as “result 2”. This will be described using
“Result 1” described above is therefore: (diff2)/2 and “result 2” is (diff2)/ (4−2). These values are of course set and stored in a prescribed work RAM region. The characteristic shown in
Next, the co-ordinates whereby the hands and/or feet etc. move when a move is executed are determined for each requested move in step 2112. These co-ordinate values are determined beforehand under the assumption that there is no difference of level with respect to the other character. This co-ordinate (leaf) is taken as “result 3”. This leaf is determined in accordance with the spatial co-ordinate position of the hand or foot when a move is executed by a function fat for returning the leaf. The processing when an attacking move is executed is described below with reference to the following and subsequent steps.
After a move has been generated, in step 2114, detection is carried out to ascertain the serial number of the frame to which the currently displayed frame corresponds. In the next step 2116, it is determined whether or not this frame is a frame prior to generation of attacking force. In the Figure, it is determined that frames up to frame 1 are frames prior to generation of attacking force.
Next, if the result of this determination in step 2116 is positive, in step 2118, this “result 1” and the frame number are multiplied and the result of this calculation is added to the leaf of the step 2112 as a difference. Using this result, the shape of the body of the warrior is then recalculated using the value etc. obtained by adding this difference to the lease1, using the known inverse kinematic technique (see for example “Dynamics And Control Of Robots”, compiled by the System Control Information Association, by T Arimoto, published by Asakura Shoten, 5.2 Reverse kinetics (page 132˜)).
In step 2116 already described, if it is determined that [the frame] is a frame subsequent to that in which attacking force is generated, in step 2200, the current frame number (“2 or 3”) in the Figure, is subtracted from the frame number (“4”, in the Figure) at which hardening is dissolved, and processing shifts to the step where this frame number is multiplied by result 2.
As a result, as shown in
With the processing that has been described, image processing is executed that reflects the difference in level of the two warriors, so an attack (punch or kick) that is performed by a single warrior is carried out to a warrior in a higher location and images matching the current situation of the actual attack can thereby be provided. Consequently, even without storing beforehand in memory motion in respect of a warrior in a higher position, image processing between one character and another character can be implemented in a condition reflecting the level difference between the characters. Incidentally, the reason why diff1 is corrected is as follows. If such correction were not carried out, when there is a large level difference between the two warriors, in an extreme case, the assault from one warrior to another warrior could be directed in the direction at right angles with respect to the ground, which would itself be unnatural.
It should be noted that, although in the explanation of the various processes described above, the example of a video game was taken, there is no restriction to this. If
In glow shading by CPU block 10 (the same applies to other types of shading), processing may be performed not only for all the polygons of a character but also for only part of a character, in particular for a portion which it is desired to show three-dimensionally by linear interpolation of colour (for example, polygons corresponding to exposed skin). In this way, the load on CPU block 10 during execution of processing can be reduced. In the case of the character of
Also, although conventionally in the case of a quadrilateral polygon four vertices were employed when Z sorting was applied to the polygons, it would be possible to determine this using only the point of intersection (mid point) of the diagonals respectively joining two vertices. As a result, processing speed can be improved by reducing the frequency with which the main CPU accesses memory. It should be noted that, even though z sorting is applied using such a mid point, the same accuracy as in the case of Z sorting using four vertices is still maintained, since this mid point represents the average value of the four vertices. In the case of a triangular polygon, the centre of gravity could be used. Next, another embodiment of processing operation of the image generating device described above will be explained.
The user i.e. the player controls a character as he wishes by operating this key and/or buttons in various manners; however, a great deal of practice is required in order to control a warrior moving very rapidly and with great versatility in an appropriate manner.
Accordingly, in a mode of the processing described herein, as shown in
Continuing the description in more detail, “P, K, P” indicates that the punch button, kick button and punch button are continuously pressed and, if we assume that an overhead throw is performed if continuation of these is determined, when P is pressed, the character goes into the punch motion. If K is pressed after the commencement of this motion but for example before the motion goes into punch return, a kick is commenced from the attitude of a delivered punch and with the timing of the punch return motion; if P is input whilst the kick motion is being executed, this is deemed as a continuous move and continuance is effected with the overhead throw motion (command move). If the timing at this point is made to coincide with the timing at which the other character is about to fall over, a very effective overhead throw can be created in which moves are easily applied to the same overhead throw. A player can create a plurality of such combination files and allocate a single button to each file. Of course, if the player presses a button to which no such combination file has been allocated, the single move that was originally allocated to this button is executed. What range of buttons are to be used to register combination moves may be suitably determined in accordance with requirements for combination moves and requirements for single movements. A plurality of buttons could be allocated to individual files.
Next, the operation of this processing will be described with reference to the flow chart shown in
In the presently described mode, a combination of action commands to a character (model) (e.g. combination of operating switch presses: pressing a single switch, pressing a plurality of switches, pressing the switches in sequence, or pressing the switches simultaneously) is stored in memory; the means for image processing reads this by a simple switch operation such as for example pressing a single button, and the movement (motion) of the model is continuously controlled in accordance with the group of commands that are thus read. Consequently, as described above, the user can represent more versatile movement (motion) of the model as he wishes, without complicated key operation.
Next, a second embodiment of a video games machine constituting an image processing device will be described with reference to
Residual Image Processing
In polygon data file #1, a model #1 corresponding to residual image 4 is stored. Model #1 of the warrior is therefore displayed as actual image on the screen at this time point. At the next time point, model #2 corresponding to residual image 3 is stored in polygon data file #2. Model #2 of the warrior is therefore displayed as actual image on the screen at this time point. Subsequently in the same way, warrior's leg models #3 . . . , #n are displayed on the screen. In the case of the fourth frame (#4) of
B#1, B#2, B#3 and B#4 are provided as buffers for the residual images. The warrior motion data recorded in the polygon data files are successively recorded in these buffers. This is displayed in the bottom portion of
For example, at the time point of the fifth frame, motion data of model #5 is stored in the polygon data file and this model is displayed as actual image. Motion data of model #4 is stored in residual image buffer #1, motion data of model #3 is stored in residual image buffer #2, motion data of model #2 is stored in residual image buffer #3, and motion data of model #1 is stored in residual image buffer #4.
Since the motion data of the polygon data file and the data of the residual image buffers (equivalent to previously calculated motion data) are displayed simultaneously, the actual image and the residual images are simultaneously displayed on the screen. This is shown in
Next, it is determined (S604) whether this attribute data is data to the effect that residual images are to be generated or the reverse. For example, in the case of “attacking move” or “move using the feet”, residual image representation is deemed to be necessary and a prescribed flag “F” is set to “1” (S606); otherwise, “F” is set to “0” (S608).
In the next step of the outline flow chart, the location where residual images are required is identified (S502). This identification is performed by the processing shown in
In
Of course, it would be possible to display (S706) residual images even if an attack command is not generated. Of course, in this case, in the determination “are residual images required?” in S500 of
Then, in S504 of
In this embodiment, the data of the polygon file having a series of movement data of a warrior are directly successively recorded in the residual image buffers. In this polygon file, there are recorded the co-ordinates of each vertex of a polygon converted to co-ordinates and clipped, together with the normal vector at each point. Since this polygon file data is directly recorded in the residual image buffers, it only needs to be output together with the current actual image, so co-ordinate conversion (modelling) of polygons for use as residual images on each occasion, as was done in the prior art, is unnecessary, thereby lightening the calculation load on the CPU block. Furthermore, since residual images are displayed in the past positions of a part (leg) and semi-transparent calculation is applied, the residual image representation can be displayed with considerable dramatic effect. It should be noted that, in addition to the semi-transparent calculation, the display of leg polygons could be represented by mesh-shapes or a line. By semi-transparent processing, the residual image of the leg and the background are displayed simultaneously.
With this embodiment, large moves, to which residual images are affixed, and medium or small moves, to which residual images are not affixed, can be distinguished by the presence/absence of residual images by the visual sense of the player, thereby improving the realism of the game (variation and interest of the image representation as a game).
Flying Material Movement Processing
Next, churning-up processing of flying material (ground material) will be described. As will be described, this processing provides a mode whereby sand or water on the ground or leaves that have fallen on the ground can fly up. A description concerning the first two of these will be given first. The gist is as follows.
As shown in
First of all, churning-up processing of sand or water will be described. In this connection, churning-up processing, as shown in
In this processing, a parts check of the warrior is performed. The warrior consists of the parts: head, right hand (upper arm, lower arm, wrist), left hand, breast, abdomen, hips, left leg (thighs, leg, ankle) and right leg. The parts check is performed for legs, head and hips. For example, [material] kicked up by the feet or falling from the head or hips is therefore simulated. The parts check includes a check of the amount of movement of the parts. Apart from this, the attribute flag of the virtual ground surface and the parts position flags are checked. The virtual ground surface attribute flag (b water) is set to “1 (water attribute)” or “0 (sand attribute)”, and the part position flag (b air) is set to “1 (grounded condition)” or “0 (aerial condition)”.
If (b water) is “1 (water attribute)”, churning-up of water is performed; if it is “0 (sand attribute)”, churning-up of sand is performed. The churning-up of water or sand comprises the following modes.
Generation on landing: this is the case where the part lands on a ground or water surface: sand or water is scattered in all directions from the point of generation.
Generation by kicking up: this occurs in the condition in which the part is grounded on a ground surface (water surface); sand or water is churned up in the direction of movement of the feet (part that is being checked) by being kicked up by the feet. Generation by kicking up and landing are shown diagrammatically in
In this processing, processing, called continuation processing, is executed, which is continued for a period of a few interrupts (a few scenes or a few frames) of occurrence of landing or occurrence of kicking-up. Usually, at the time of such occurrence, the churning-up number is determined and set to ¼ thereof. If the churning-up number is taken to be N, the churning-up number is successively decreased as follows with each interrupt.
If the churning-up number is taken to be N:
That is, it is set to ¼ in each case of the remainder of the churning-up number at each interrupt. Regarding the mode of churning-up, a pattern change is arranged to be performed as already described.
Processing in the case of landing occurrence will now be described. Let the position of the checked part on the previous occasion (i.e. one interrupt previous) be (OX, Oy, OZ), and the present position be (PX, Py, PZ). Let the height of the ground surface or water surface (Y co-ordinates) be epos.
Basic speed of the part:
Total number of churning-up (set number): Amount=Basic set number. Part movement amount.
Point of origin of churning-up (point of origin of occurrence):
Processing of dispersion of sand (water drops): the X and Z components of the basic speed of the checked part are extracted.
Let us now take A as a vector that is randomly rotated. Next, take A′ as being the result obtained by moving L in the direction of the vector along this vector. In this case “L” is the radius of the part. For example, in the case of a leg this is set to 10 cm, in the case of hips this is set to 20 cm, and in the case of the head this is set to 15 cm. The final speed of churning-up of sand and water droplets is determined by adding this A′ to the Y component of the basic speed. After this speed has been determined, the sand or water droplets etc. are moved along the track shown in
Next, the processing of occurrence of churning-up will be described. Let us assume that a foot is kicked upwards in water or sand as shown in
The reason for finding COS θ in this case is that the amount of churning-up decreases as the angle of churning-up gets larger, as it gets closer to (90°). That is, the set number for the water/sand is: Basic set number. Amount of movement. COS θ.
The speed of the sand/water drops is a value obtained by multiplying the basic speed by the numerical value of a random number (0˜255)/2550. This is identical with the previous case of occurrence on landing. Variations in sand/water scattering occurring randomly can be represented by multiplying by a random number.
The point of generation of scattering of sand or water is obtained by further adding (OX, Oy, OZ) to this speed. Referring to
Continuation processing, as described above, is used to effect continuous display of sand/water scattering, using the landing occurrence point or churning-up occurrence point that has thus been determined.
Depending on this determination, if the checked part is below the water surface, processing shifts to processing of occurrence of landing, while, in the opposite case to this, it shifts to processing of occurrence of churning up.
In S1606, a determination is made as to whether or not [the character] was in the grounded condition at the previous interrupt, by checking the landing determination flag. If this flag (b air) is “0”, this grounded condition is continued for a few interrupts (S1608). On the other hand, if this flag is “1”, it is concluded that landing has just occurred at the present interrupt, and landing occurrence processing (S1610) is performed; also, this flag is reset to “0” (S1612) and return is effected to the main routine of
However, if the Y co-ordinate of the checked part is higher than the position of the water surface, processing shifts to S1614 and [a determination] is performed as to whether or not the character was in the grounded condition on the preceding interrupt, by checking the landing determination flag. If, on the previous occasion, the character was in a landed condition i.e. the checked part was below the water surface, it is assumed that, for example, kicking-up by the feet as shown in the Figure has currently taken place and the upwards churning-up pattern is displayed (S1616) matching the movement of the feet through the water. This flag is then set to “1” and return (S1617) is executed. On the other hand, if the character was not in the landed condition on the previous occasion, churning-up processing is continued (S1618).
As described above, with this embodiment, the amount of movement and the direction of movement of a character can be reflected by the surface material (water or sand etc.) that is moved to match the movement of the character.
Next, processing will be described that simulates flying-up movement (one type of flying material movement processing) of leaves (one type of flying material) that has fallen on to the stage caused by a wind generated by movement of a character (warrior) in the game.
Referring to
This processing is executed in accordance with the diagrammatic flowchart shown in
First of all, a flow chart for checking generation of wind will be described with reference to
The wind generation position and the vector of its X, Y, Z direction are as shown in the Figure. M′ is corrected to 0.12≦M′≦1.0. This M′ is the vector of the Y direction of the wind. The vectors of the X and Z directions are taken to be values obtained by multiplying the amounts of change of the X direction and Y direction by 0.6. The reason for multiplying by 0.6 is that the vectors in the X and Z directions are taken to be relatively smaller in value than the vector in the Y direction, so as chiefly to simulate the condition of wind blowing upwards in model space.
Wind is deemed to generate only a single vector during a single interrupt. If wind is generated in more than one location, wind is deemed to be generated in whichever of these has the largest value.
Next, a flow chart of the check for blowing-up of leaves will be described (
In step 2000, the wind data and leaf data are read. These data (X component, Y component and Z component) are as shown in the Figure. Calculation regarding the height of the leaf is performed in S2002. This calculation is as follows. Letting the height from the ground be epos, a height 1 m above the ground is epos +1.0 (m). Let us take C1 as the value obtained by subtracting the height of the leaf 1 ypos from this height.
C1=epos+1.0−1 ypos(m)
If C1≦0, the position of the leaf (more than 1 m above the ground) is deemed to be sufficiently high as not to be effected by the wind produced by actions of the character, so processing is terminated. Under these conditions, the leaf is not effected by wind (degree of effect 0%). On the other hand, if it is determined that there will be an effect due to wind (S2004: No) i.e. if C1>0, the effect coefficient corresponding to the value of C1 is determined (S2006). For example, if the leaf is on the ground, (1 ypos=epos, C1=1) so the leaf undergoes 100% of the effect of the wind. If C1=0.5, it undergoes 50% of the effect. C1 takes a value 0 s C1<1.0. This is explained in
If w yspd −L′ is 0 or less, the leaf is deemed to be outside the range of effect of the wind, so there is deemed to be no effect from the wind and return is executed (S2010). On the other hand, if it is more than 0, C2 (effect coefficient dependent on distance) is calculated. Further, this C2 is corrected to calculate C2′. An example of this effect coefficient is shown in
In S2014, the effect of the wind on the leaf (vector: X direction, Y direction, Z direction) i.e. the speed of the leaf (vector) is determined as shown in the Figure; thus the movement of the leaf is simulated (S1806 of
As shown in
(1) shows movement in which the centre position of the leaf is displaced by the effect of wind or by weight and (2) shows fluttering movement of the leaf simulated by X rot and Y rot. These movements are described by a diagrammatic flow chart and constitute one link of pre-fall movement.
A characteristic expression for the former movement (when in the air and when landed from the air) is as shown in S2100 of the Figure. The reason for multiplying the speed in the X direction (1 xspd) by 0.87 to obtain the new speed in the X direction in the case where the leaf is in the air is to take into account air resistance. The same applies in regard to the Z direction. The new speed in the Y direction is obtained by subtracting 0.002 from the speed in the Y direction. This simulates gradual increase in the speed in the direction of the weight of the leaf. However, the maximum speed of fall in the Y direction (1 yspd) is taken as −0.006. The case where the leaf has landed will be described later.
In contrast, as shown in
Y rotational speed=1 yspd×1200×256
One-turn rotation (360°) is expressed by 0×0000˜0×FFFF (0˜65536). This is 1° (solid black block symbol)2 182 (0×b6). It should be noted that, if Y rot were also subjected to pattern changing, there would be a risk of leaf movement simulation appearing to be coarse, so only one rotational element (X rot) is subjected to pattern change, the remaining element (Y rot) being found by calculation.
As shown in
Whether a leaf that had been floating in the air has landed or not is determined by comparing the Y co-ordinate of the ground surface and the Y co-ordinate of the leaf. If the Y co-ordinate of the leaf ≦ (or “=”) the Y co-ordinate of the ground surface, the leaf is determined to be in a grounded condition. Shape modelling of the leaf is effected by mapping a leaf design (texture) on to a single polygon.
Since the amount of movement and direction of movement of the flying material (water/sand/leaf) as described above is determined by the movement characteristic of the character (model), the movement of the model can be precisely reflected in the flying material. Such flying material is made to fly up or falls down in response to the movement of the original model, so the actions of the flying material can be simulated with high quality by reflecting the movement of the model to these.
Also, with the movement processing of moving material described above, (free) falling movement of the model (flying material) in virtual space can be reproduced with a high degree of dramatic effect and with high quality. For example, images of fluttering leaves can be generated.
Free-Fall Movement Processing
Next, movement processing of an object falling in virtual space will be described. In this embodiment, the object is taken to be snow, and simulation of dramatic effects by applying a “swirling” movement to the snow will be described. In this connection, what is meant by “swirling” is a comparatively irregular movement such as a condition in which snow falls whilst being blown about by the wind or a condition in which the snow rises and falls in the manner of a falling leaf after being stirred up by movement of a warrior. This embodiment efficiently simulates this “swirling” condition. A practical example of the swirling of snow will now be described. The swirling and falling snow is constituted by one or more polygons and is affected by the wind vector that is set up in model space (this vector may be suitably varied). The movement of the snow polygons is simulated by the effect produced by free fall due to its weight and the wind vector (the foregoing is the first means of the claim) and by circular movement (second means of the same claim) in the XZ plane of virtual space (horizontal plane i.e. co-ordinate system parallel to the ground surface). The angular speed and amplitude of the circular movement are stored in tabular form in the memory and one of these is selected for use.
The circular movement described above that applies swirling to the behaviour of the snow is applied as follows. Taking the angular speed as ω, the angle of rotation as q and the amplitude as a, the amount of offset with respect to the X component (x off) and Z component (z off) of the swirling component i.e. the basic behaviour of the snow (determined by free fall and a vector due to the wind) is as follows.
x off=sin(q+ω)·a
y off=cos(q+ω)·a
This offset data is calculated ten times during each interrupt and added to the basic position of the snow.
The equation of motion of the snow is as follows.
In the present embodiment, the swirling and falling behaviour of snow blown about by the wind (see
By means of this processing, the quality in simulating free-fall movement of a model in virtual space can be raised, so the dramatic effect of free-fall image processing can be manifested to a high degree.
Collision determination processing in irregularly shaped zone Next, collision processing between an irregularly shaped zone and a model will be described. This processing has the following significance. In this connection, a zone is a region restricting movement of a warrior and can easily be understood by for example imagining a ring whose periphery is enclosed by a wall of irregular shape. The warrior can move freely within this ring but cannot cross the wall to move outside the ring. This wall corresponds to a “zone” in this description.
An irregularly shaped zone 250 is shown at (A) of
As shown in (A), in this collision determination, the shape of the character is approximated by a sphere; furthermore, the wall and this sphere are projected in parallel on to a plane (XZ plane) parallel to the ground surface so as to find the positional relationship between the sides forming the region of the zone and the circle 252 obtained by projecting the sphere. The position of the sphere is the calculated position of the character. In this Figure, L is one vertex of an arbitrary side of this zone and R is another vertex of this side. This is set such that L comes in the anticlockwise direction with respect to an arbitrary side. T in the Figure is the centre-point of the projected circle of the character. Vector R, vector T, and vector L are vectors set from some arbitrary point of this co-ordinate system to these various points. Also, vector P is a vector drawn perpendicularly from the centre co-ordinate of the circle to each side of the zone. Vector P is a vector drawn at right angles to side LR from the point of intersection of the circle and the tangent to this circle parallel to side LR, and corresponds to the expulsion vector in this process. The object of this process is the calculation of expulsion vector V. When the centre of the circle is outside straight line RL (this is assumed to be a positive region), as shown in the Figure, an expulsion vector is set up as shown by expression (1) in the Figure. In contrast, when the centre of the circle is within straight line RL (this is taken to be the negative region), it is set up as in expression (2). In other words, if the distance between the centre of the circle and straight line RL is more than radius (d), an expulsion vector is not set up. If this distance is exceeded, an expulsion vector is set up. In this case, since the vector is directed in the direction of the wall, unit vector Pe is set up in the opposite direction to when the centre of the circle is outside straight line RL.
Next, in S2604, a comparison is made of this character position and a given side RL. Specifically, the relationship of expression (3) of
If this relationship is affirmed, the value of vector P is calculated in accordance with what is shown in
In the next step (S2614), a determination is made as to whether or not the centre of the circle is within the zone: if it is outside the zone, an expulsion vector is setup in accordance with expression (1) of the Figure; if it is within the zone, an expulsion vector is set up in accordance with expression (2). In the case of expression (1) and (2), if the magnitude of vector P is smaller than the radius of the circle, a collision is deemed to have occurred between the character and the wall and the expulsion vector described above is set up. By means of this expulsion vector, image processing is executed (S2616) whereby further movement of the character beyond this zone (wall) is restrained in the actual image.
Of course, image processing could be executed whereby, by the expulsion vector from this wall, the model is pulled by vector V defined in the normal direction of the side from the wall.
With the processing indicated in this embodiment, an expulsion vector is arranged to be calculated for each side, so, even in the case of an irregularly shaped zone, determination of collision between this zone and the character can be carried out in a reliable manner (specifically, collision determination is affirmed when the value of the expulsion vector is given) and the results of this determination can be reflected in image processing.
Compilation of Texture Series
Next, a mode of compiling a repeated texture series will be described.
The third texture series is a texture series wherein an image (pattern of reflection on sea) which is an image repeated with a prescribed period is formed at a target.
The third texture series is constituted of 30 still pictures 0˜29; these pictures are successively mapped onto a polygon (i.e. the pictures are mapped onto polygons 0˜29), then further mapped onto numbers 0˜29; by repeating these operations, it is arranged that the picture changes at the joining portions of these texture series (for example 28˜2) take place in a naturalistic manner (continuously, in other words, without skipping pictures). By this means, an image in which the reflection pattern on the sea surface is repeated can be compiled.
Thus the first and second texture series are compiled using commercially available application software, as already described. In the generation of the texture series, the parameter values are different for respective series. However, due to software limitations, the texture groups of the first region and last region of the texture series are designs showing poor linkage, though they are the same in respective series. Specifically, these are for example number 0 and number 29 of the first texture series and number 0 and number 59 of the second texture series.
Accordingly, in this processing, a transparency parameter that successively changes (0˜100%) is applied to the second texture. The numerical values above the second texture series are the values relating to this transparency. 100% indicates a condition in which there is no transparency at all; apart from this, 7% indicates a condition in which transparency is more advanced (93% transparency). 0% gives complete transparency. A parameter of 100% is applied to the first texture.
The third texture series is compiled by successively superimposing the first texture series and the second texture series. Specifically, number 0 of the third texture series is formed by superimposing number 0 of the first texture series and number 30 of the second texture series, number 1 of the third texture series is formed by superimposing number 1 of the first texture series and number 31 of the second texture series, number 29 of the third texture series is formed by superimposing number 29 of the first texture series and number 59 of the second texture series.
In this process, when number 0 of the first texture series and number 30 of the second texture series are superimposed, since the transparency of number 30 of the second texture series is 0 (completely non-transparent condition), number 0 of the first texture series is completely hidden by number 30 of the second texture series so the image of number 0 of the third texture series is equivalent to the image of number 30 of the second texture series. In contrast, when number 29 of the first texture series and number 59 of the second texture series are superimposed, since the transparency of number 30 of the second texture series is 97% (practically transparent condition), the image of number 29 of the third texture series is practically equivalent to the image of number 29 of the first texture series (in other words, the image of number 59 of the second texture series is practically invisible). Thus, when the second texture series is generated, the method of applying the parameter is such that the picture of number 30 is a picture that continues from the picture number 29 of the first texture series, so, when the third texture series is viewed, a picture can be generated whose joining is naturalistic, as described, at the joining section (number 28˜number 2) of the texture series. By assigning parameters to the respective texture series such that the picture of the end section of the first texture series (for example number 25 to 29) and that at the initial section of the second texture series (for example number 30 to number 34) link up (or effectively do so), as shown by the arrow in the Figure, by taking into account transparency, continuity can be achieved from number 29 of the first texture series to number 30 of the second texture series, thereby producing pictures in a naturalistically linked mode as illustrated in the drawing by the third texture series.
By storing the third texture series in a prescribed region of memory and successively mapping this onto the polygons indicating the sea surface, images can be displayed in which the pattern and shape of white waves on the sea surface (portions of high brightness on the screen i.e. portions that are displayed as white in
The speed with which the wave mode is reproduced and the number of textures employed can be altered as required. Taking the case of a stage in which a warrior is standing at the centre of the screen, in the near area of the stage (i.e. the vicinity of the centre of the screen), a reproduction speed of ( 1/30 sec per frame, 30 frame reproduction) as described may be assumed; for waves further than this, in order to reduce the calculation load on the computer, the reproduction speed per frame can be lowered and the total number of frames used can be reduced. For example, frames number 0, number 10, number 20 and number 29 of the third texture series may be employed. The reason for doing this is because, even though the reproduction of the images in the vicinity of the periphery of the screen is somewhat rough, the adverse impression on the player is slight and there is more benefit in the lowering of the calculation load on the computer graphics device.
The third texture series is stored in memory of the games device as described. A character series as described above can therefore by utilised in the field of games machines employing computer graphics. The described method of generating a character series can be used not only in the technical field of computer graphics but also in the technical field of compilation of animation video.
Slow Reproduction Processing
In contrast, as shown in
Thus, let us now suppose that one character (warrior) performs a move (“punch”) on another warrior. On the other hand, let us assume that the other warrior practically simultaneously performs a move parrying this punch. It should be noted that this punch or parry could be generated by suitable operation of a control button and/or direction key by the player.
When parrying of this punch is established, a motion is generated whereby the attacking warrior recoils considerably to the rear. At this point, the speed of reproduction of this motion is made smaller than the speed of reproduction of the images of the defending warrior. When this happens, due to this slow reproduction, an “opening” is produced in respect of the attacking warrior, making it easy for the defending warrior to perform an attacking move on the attacking warrior. Once warrior 292 has recoiled, he gradually returns to his original stance (before recoil) as in
In S3412, the amount of parry is divided by the set number of frames and a leaf amount (amount of movement and direction of movement in the spatial co-ordinate system) corresponding to one frame is calculated. Then, in S3414, the speed of reproduction (display speed) of the attacking warrior is made for example half that of the defending warrior. For example, the reproduction speed of the attacking character is made 1/30 sec, half of 1/60 sec. Next, in S3416, the timer value is determined;. if this is “0”, slow reproduction is deemed to have been completed and the reproduction speed of the attacking warrior is returned to the original condition (S3417) and return is executed. On the other hand, if it is not “0”, the leaf amount for one frame is multiplied by the timer value and this is regarded as an amount of movement by the attacker. Then, in S3420, this amount of movement is added to the co-ordinate value of the attacking warrior. Specifically, in the position of attacking warrior 292 of
Next, in S3422, 1 is subtracted from the value (T) of the timer, and, in S3424, the motion position of the attacking warrior 292 (
With this processing, if the parry succeeded, images are reproduced wherein the attacking warrior, after first of all recoiling considerably, gradually recoils further in the slow reproduction condition. At this point, the defending warrior can easily perform an attacking move on the attacking warrior.
With this processing, reproduction of motion by another warrior can be performed whilst the reproduction speed for motion by a first warrior is lowered, so the player can more effectively perform fighting moves of the defending warrior that he is controlling: this has led to the provision of a facility for slow reproduction as a game element. Slow reproduction can therefore enable a high degree of dramatic effect to be achieved.
It should be noted that the present invention is not restricted to the embodiments described above but could be further modified in various ways by persons skilled in the art within the scope of the claims.
As the memory (recording) medium for storing the operating program of the games machine, apart from cartridge ROM or CD-ROM as described above, communication media such as the Internet or personal computer networks could be used; an electronic mail server is also included.
| Number | Date | Country | Kind |
|---|---|---|---|
| 8-165224 | Jun 1996 | JP | national |
| 8-186666 | Jun 1996 | JP | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 09962594 | Sep 2001 | US |
| Child | 11173057 | Jul 2005 | US |
| Parent | 09011023 | Aug 1998 | US |
| Child | 09962594 | Sep 2001 | US |
