Patent Application
20090046098
1. Field of the Invention
The present invention relates to the technical field of graphic rendering and, in particular, to a primitive binning method for use in a tile-based rendering system, for example in a sort-middle technique.
2. Description of the Related Art
A virtual three dimensional (3D) model (or simply “3D model”) is comprised of primitives in the form of polygons, such as triangles, which represent the skin of the 3D model. A graphic 3D engine draws polygons from the 3D model onto a two-dimensional (2D) surface, such as a screen.
A summary of the prior art rendering process can be found in: “Fundamentals of Three-Dimensional Computer Graphics”, by Watt, Chapter 5: The Rendering Process, pages 97 to 113, published by Addison-Wesley Publishing Company, Reading, Mass., 1989, reprinted 1991, ISBN 0-201-15442-0.
In a traditional pipeline, the primitives are processed in a submission order. A more efficient method is to break up the frame buffer into individual subsections (tiles) and to render them individually. Each tile includes one or more polygons or, more typically, a portion of one or more polygons.
A tile based rendering employs therefore a way to associate tiles covered by a primitive. A rasteriser renders all primitives of one tile, so which tile is covered by that primitive is found first.
To reduce the amount of tiles that each polygon is assigned to, a primitive or polygon binning method may be used. A polygon binning method excludes tiles that do not include any polygons or portions thereof prior to rasterization. The binning process also accomplishes some rasterization setup by identifying which polygons are contained by each tile.
A simple binning method provides for constructing a bounding box around the primitive. However, many tiles of the bounding box may still be outside the primitive.
According to another approach, the binning method calculates the equations of the lines formed by the edges of the primitive (in the standard y=mx+c format) and then tracks up the lines from vertex to vertex enabling the tiles which are transversed. This (depending on the size of the triangle) may leave some tiles in the middle. These can be checked and enabled by traveling along each row and enabling all tiles which are situated between two previously enabled tiles. This method requires to calculate reciprocal values and has therefore a very high cost.
According to an alternative method, the equations of the lines which form the sides of a polygon are not used and edge equations, which can be derived from a matrix formed by the vertex co-ordinates (“outcode method”), are employed. Therefore, each side of the triangle is associated with an edge equation. Any point on the line will satisfy this equation, with points on one side giving a positive evaluation and on the other a negative result. This property can be used to determine on which side of a line a point is, and therefore, by using three edge equations, whether a point is inside a triangle or not. In the present case, a tile is considered to be covered by a primitive if at least one of its corners is within the triangle. Therefore, this method requires checking against three edge equations for all the corners of the tiles. If all four corners of any tile are outside any of three edges, then that tile can be ignored.
According to another approach, called “midpoint method”, triangles are tiled by finding the midpoint of the edges and then disabling blocks of tiles that are outside these points. If a high accuracy is desired, also this method turns out to be quite cumbersome.
There is therefore the need of associating tiles to primitives in a more efficient way, in order to reduce the amount of processing to be performed by the 3D graphic engine.
One embodiment of the present invention provides a primitive binning method that includes detecting border tiles of a primitive defined by at least three vertexes. The detecting includes:
defining a left edge and a right edge of the primitive compared to a direction of exploring tiles;
calculating a slope sign for the left edge using an edge equation for the left edge and calculating a slope sign for the right edge using an edge equation for the right edge; and
checking if a tile is crossed by one of the edges by evaluating an edge equation of a single corner of a tile, the corner being selected according to the one of the edges being a left or a right edge and according to the slope sign of the one of the edges.
The features and advantages of the method according to various embodiment of the invention will be apparent from the description given below, with reference to the following figures, in which:
As an example, the mobile phone 100 can be a cellular phone provided with an antenna 10, a transceiver 20 (Tx/Rx) connected with the antenna 10, an audio circuit unit 30 (AV-CIRC) connected with the transceiver 20. A speaker 40 and a microphone 90 are connected with the audio circuit unit 30.
The mobile phone 100 is further provided with a CPU (Central Processing Unit) 60 for controlling various functions and, particularly, the operation of the transceiver 20 and the audio circuit unit 30 according to a control program stored in a system memory 80 (MEM), connected to the CPU 60. Graphic module 500 is coupled to and controlled by the CPU 60. Moreover, mobile phone 100 is provided with a display unit 70 provided with a corresponding screen 71 (e.g., a liquid crystal display, DSPY), and a user interface 50, such as an alphanumeric keyboard (K-B).
The graphic module 500 is configured to perform a set of graphic functions to render an image on the screen 71 of the display 70. In one embodiment, the graphic module 500 is a graphic engine configured to render images, offloading the CPU 60 from performing such tasks. In one embodiment of the present invention the term “graphic engine” means a device which performs rendering in hardware. The terms “graphic accelerator”, also employed in the field, is equivalent to the term graphic engine.
Alternatively, the graphic module 500 can be a graphic processing unit (GPU) wherein the rendering functions are performed on the basis of hardware and software instructions. In accordance with a further embodiment, some or all of the rendering functions are performed by the CPU 60.
In
In accordance with the sort-middle rendering, the screen 71 of the display 70 is divided in a plurality of 2D (two dimensional) ordered portions (i.e., 2D tiles) such as, for example, square tiles 5 as shown in
The graphic engine 500, illustrated in
The driver 501 is a block having interface tasks and is configured to accept commands from programs (e.g., application protocol interface—API) running on the CPU 60 and then translate them into specialized commands for the other blocks of the graphic engine 500.
The geometry stage 502 is configured to process primitives and apply to them transformations so as to move 3D objects. As defined above, a primitive is a simple geometric entity such as, e.g., a point, a line, a triangle, a square, a polygon or high-order surface. In the following reference will be made to triangles, which can be univocally defined by the coordinates of their vertexes, without other types of employable primitives.
The binner stage 503 is adapted to acquire from the geometry stage 502 primitive coordinates and associate them to each tile of the screen 71. The binner stage 503 is coupled to a scene buffer 504 which is a memory able to store information provided by the binner stage 503. As an example, the scene buffer 504 is a memory external to the graphic module 500 and can be the memory system 80 illustrated in
The graphic module 500 further includes a parser stage 506, a rasterizer stage 507 and a fragment processor 508 which is coupled to the display 70. The parser stage 506 is responsible for reading, for each tile, the information stored in the scene buffer 504 and passing such information to the following stages also performing a primitive reordering operation.
The parser stage 506 generates an ordered display list which is stored, temporarily, in a parser side memory 509. The parser stage 506 is suitably coupled to the scene buffer memory 504 in order to read its content and is coupled to the binner stage 503 to receive synchronization signals.
According to one embodiment, the parser side memory 509 may be an on-chip memory, which allows a fast processing. As an example, the parser side memory 509 is integrated on the same chip on which the parser stage 506 has been integrated and, e.g., shows a capacity of 8 kB.
The rasterizer stage 507 is configured to perform processing of primitive data received from the parser stage 506 so as to generate pixel information images such as the attribute values of each pixel. The attributes are data (color, coordinates position, texture coordinate etc.) associated to a primitive. As an example, a triangle vertex has the following attributes: color, position, coordinates associated to texture. As is known to the skilled person, a texture is an image (e.g., a bitmap image) that could be mapped on the primitive.
The fragment processor 508 defines fragments from the received pixels, by associating a fragment depth to pixels and other data and performing suitable tests on the received pixels.
Moreover, the particular geometry stage 502 described comprises: a lighting stage 551, a primitive assembly stage 552, a clipping stage 553, a “perspective divide” stage 554, a viewport transformation stage 555 and a culling stage 556.
The per-vertex lighting stage 551 applies light to the primitives depending on a defined light source and suitably adjusts the primitive color vertexes in such a way to define the effect of the light. The primitive assembly stage 552 is a stage that allows reconstruction of the semantic meaning of a primitive so as to specify the primitive type, i.e., specifying if a primitive is a triangle, a line or a point and so on.
The clipping stage 553 allows removal of the primitives that are outside the screen 71 (non-visible primitives) and converting the primitives that are placed partially out of the screen 71 into primitive which are fully visible. The perspective divide stage 554 is adapted to apply a projective transformation dividing each coordinate value by a vector w.
The viewport transformation stage 555 is configured to apply a further coordinates transformation which takes into account the screen resolution. The culling stage 556 has the task of removing the primitives oriented in a direction opposite to the observer and its operation is based on a normal direction associated to each primitive.
In operation, the user of the mobile phone 100 employs the keyboard 50 in order to select a 3D graphic application, such as a video game. As an example, such graphic application allows to show on the screen 71 several scenes. The scenes correspond to what is visible for an observer who can move assuming different positions. Accordingly, a software module corresponding to said graphic application runs on the CPU 60 and activates the graphic module 500.
A 3D scene to be rendered is included in a region of space, called view frustum VF (
The binner stage 503 associates empty tiles with the triangle to avoid redundant rasterizer calculations. It is clear that, if triangles are smaller then tiles, the binner stage 503 processes all triangles within each tile before proceeding to the next tile. If the triangles are larger than tiles, it associates the triangles with all the tiles they cross and stores the state. In this case, an exploration of the tiles is carried out.
According to one embodiment, the binner module 503 is adapted to detect the tiles crossed by the edges of a triangle (border tiles), as described later in more detail. All the tiles between two border tiles on the same row are then considered included in the primitive and may therefore be stored.
Method 600 is directed to detect the tiles covered by a primitive in the form of a triangle 6 defined by three vertexes. It has to be noted, however, that the method here described is also applicable to other polygons, since it is always possible to decompose a polygon in triangles.
According to one embodiment, before starting the exploration of the tiles, the binner module 503 defines, by means of computations, a bounding box 7 around the triangle 6 (step 601 and
In a step 602, the binner module 503 performs a primitive set up phase, in which for each couple of vertexes the equation of the line passing through the vertexes is calculated in the form of the following edge equation:
E=x(y1−y0)−y(x1−x0)−x0(y1−y0)−y0(x1−x0)=ax+by+c
Any point on the line will satisfy this equation; points not belonging to the line and placed on one side will give a positive result, while points not belonging to the line and placed on the other side will give a negative result. Therefore, the edge equation can be used to determine on which side of a line a point is placed.
The three edge equations of the triangle will be:
E0=x(y1−y0)−y(x1−x0)−x0(y1−y0)−y0(x1−x0)=a01x+b01y+c01
E1=x(y2−y0)−y(x2−x0)−x0(y2−y0)−y0(x2−x0)=a02x+b02y+c02
E2=x(y1−y2)−y(x1−x2)−x2(y1−y2)−y2(x1−x2)=a21x+b21y+c21
In the set up phase, the binner module 503 defines if an edge is left or right and the sign of the slope of each edge. According to this information, the binner module 503 selects, for each edge, the tiles scan direction and just one corner of the tile to be evaluated. In fact, given a scan direction and the slope of an edge, it is sufficient to test just a corner of a tile to check if the tile is crossed by the edge.
More in detail, the following four combinations can occur for an edge 8 (
a. Left edge and positive slope;
b. Left edge and negative slope;
c. Right edge and positive slope; and
d. Right edge and negative slope.
In the set up phase the binner module 503 performs the following selection:
If an edge 8 is left and has a positive slope (case a), the tiles scan direction is from left to right and the tile corner to be evaluated is the bottom-right (
if an edge 8 is right and has a negative slope (case b), the tiles scan direction is from right to left and the tile corner to be evaluated is the bottom-left (
if an edge 8 is left and has a negative slope (case c), the tiles scan direction is from right to left and the tile corner to be evaluated is the top-right (
if an edge is right and has a positive slope (case d), the tiles scan direction is from left to right and the tile corner to be evaluated is the top-left (
According to one embodiment of the invention, for determining if an edge is left or right, the binner module 503:
As an example shown in
indexMiddle=3−indexMaxY−indexMinY.
As stated above, to verify if a point is inside or outside a primitive, the sign of the edge equation may be tested. The line equation could be ax+by+c=0 or −ax−by−c=0, so the sign of the edge equation depends on the sign of the coefficients. According to one embodiment, it is better to have always an internal point with a positive edge equation for all edges, so it is not necessary to change sign of edge equation every time. In other words, the coefficients sign of the edge equation of the left and right edges is chosen such that the results of the edge equations for a point falling between the left and right edges have the same sign.
Then it is sufficient to follow a clockwise direction to define the sign of coefficient.
With reference to
If E0(V2)<0, then the middle vertex is left and the sign of a01, b01, c01 may be changed to have an internal point with E0, E1, E2 positive (
A winding equation (cross product between two vectors) gives the same result as the edge equation, so they are equivalent.
According to one embodiment, the set up phase 602 of the method also includes a splitting of the triangle 6 into a top sub-triangle 6′ and a bottom sub-triangle 6″. In more detail, a line E3 parallel to the tiles scan direction, horizontal in the example here described, is drawn from the middle vertex V2 towards the reference edge E0. In this way, in each sub-triangle the left and right edges have the same Δy.
Having determined, for each edge, the scan direction and the tile corner to be evaluated, the binner module 503 can start the exploration of tiles to discover which are crossed by the edges (step 603).
It has to be noted that left and right edges and edge slope sign are calculated once in the set up phase for the selection of the scan direction and of the tile corner. During the scan line of an edge, the direction and the tile corner is always the same.
As stated above, according to one embodiment, the tiles exploration process is not a full exploration of the bounding box, but an exploration along a scan line defined by an edge. At the beginning of the process a first row may be explored, for example the row including the vertex with the minimum y coordinate, starting from an end (step 604). The process ends after determining that the previous tile evaluated was the last tile (step 608).
When a border tile is detected, its position is used for determining the starting point for the tiles exploration along an adjacent row.
If, instead, the tile corner to be evaluated is a top corner (in
In any case, except for the first row, when the algorithm changes row it is not necessary to restart from the first tile of the row. Considering the exploration made by row, the edge equation is calculated in an incremental way except for start row tiles, such that E new=E old+K.
It means only one sum instead of the 2 sums and 2 multiplications of the edge equation in the usual extended form: E=a*x+b*y+c.
When a tile is found to be crossed by an edge, it is marked as a start tile 5′ (if the edge is left) or as an end tile 5″ (if the edge is right). When all the edges have been explored, all the tiles comprised by a start tile and an end tile of a row are considered to be covered by the primitive and therefore are stored in the scene buffer (
It is evident from the description above that the method of finding border tiles according to one embodiment of the invention does not require to evaluate the edge equation of all tile corners and of all three edges, but of just one corner and just one edge. Therefore, the method here proposed is 12 times less complex.
Advantageously, it is possible to use this technique also in the case represented in
Another advantage of the method according to one embodiment of the invention is that it can be also used to find border pixels of a primitive into rasterizer stage 507, for example starting edge equation evaluation from the center of the pixel.
The various embodiments described above can be combined to provide further embodiments. All of the U.S. patents, U.S. patent application publications, U.S. patent applications, foreign patents, foreign patent applications and non-patent publications referred to in this specification and/or listed in the Application Data Sheet, are incorporated herein by reference, in their entirety. Aspects of the embodiments can be modified, if necessary to employ concepts of the various patents, applications and publications to provide yet further embodiments.
These and other changes can be made to the embodiments in light of the above-detailed description. In general, in the following claims, the terms used should not be construed to limit the claims to the specific embodiments disclosed in the specification and the claims, but should be construed to include all possible embodiments along with the full scope of equivalents to which such claims are entitled. Accordingly, the claims are not limited by the disclosure.
