This application is a continuation of PCT International Application No. PCT/JP2005/002467 which has an International filing date of Feb. 17, 2005, which designated the United States of America.
1. Technical Field
The present invention relates to an image processing method for continuously taking an image of an image pickup object and detecting the range of the image pickup object in an image based on image pickup; an image processing system to which the image processing method is applied; an image pickup device and an image processing device which are used in the image processing system; and a computer program for realizing the image processing device, and in particular to an image processing method, an image processing system, an image pickup device, an image processing device and a computer program for facilitating a process such as detection of a range.
2. Description of the Related Art
Suggested as a device to support driving of a vehicle such as a car is an image processing device for taking an image of the face of the driver with an in-vehicle camera, which is arranged at a location capable of taking an image of the face of the driver, and performing image processing to detect the facial contour of the driver, the location of the eyes and nose and also the status of the face based thereon from an obtained image (see Japanese Patent Application Laid-Open No. 2004-234367, for example). Moreover, constructed using such a device is a system for detecting the status of the driver and providing driving support such as warning to inattentive driving.
However, when the face of the driver is away from the in-vehicle camera, for example, there is a problem that the part of the face in the taken image becomes small and image processing such as detection of the contour and the location of the eyes and nose cannot be performed sufficiently.
The present invention has been made in view of such circumstances, and it is an object thereof to provide: an image processing method for performing manipulation such as enlargement for a second image, which is taken after a first image, on the basis of the range of the facial contour or the like detected for the first image in continuously taking an image of an image pickup object such as the face of the driver and detecting the range of the image pickup object in the manipulated second image so as to facilitate image processing such as detection of a range related to an image object; an image processing system to which the image processing method is applied; an image pickup device and an image processing device which are used in the image processing system; and a computer program for realizing the image processing device.
An image processing method according to the first aspect is an image processing method for continuously taking an image of an image pickup object and detecting a range of the image pickup object in an image based on image pickup, characterized by comprising the steps of: manipulating a second image after a first image on the basis of a range of the image pickup object detected from the first image; and detecting a range of the image pickup object in the manipulated second image.
An image processing system according to the second aspect is an image processing system comprising image pickup means for continuously taking an image of an image pickup object and image processing means for detecting a range of the image pickup object in an image based on image pickup of the image pickup means, characterized by comprising: manipulating means for manipulating a second image after a first image on the basis of a range of the image pickup object detected from the first image; and means for detecting a range of the image pickup object in the manipulated second image.
An image processing system according to the third aspect is the one according to the second aspect, characterized in that the manipulating means is constructed to enlarge the second image.
An image processing system according to the fourth aspect is the one according to the third aspect, characterized in that the manipulating means is constructed to decide an enlargement ratio on the basis of a size of a range.
An image processing system according to the fifth aspect is the one according to the third aspect or the fourth aspect, characterized by further comprising means for deciding a reference point of manipulation in the image pickup object, wherein the manipulating means is constructed to enlarge an image with the reference point being a center.
An image processing system according to the sixth aspect is the one according to the fifth aspect, characterized in that the image pickup object is a face, the range of the image pickup object is a range of the face in a predetermined direction, and the reference point in the image pickup object is a point obtained based on location of eyes and/or a nose.
An image pickup device according to the seventh aspect is an image pickup device, which comprises means for continuously taking an image of an image pickup object and is connected with an image processing device for detecting a range of the image pickup object in an image based on image pickup, characterized by comprising: means for accepting input of range data, which indicates a range of the image pickup object detected from a first image, from the image processing device; means for manipulating a second image after the first image on the basis of the range indicated by the accepted range data; and means for outputting image data, which indicates the manipulated second image, to the image processing device.
An image processing device according to the eighth aspect is an image processing device for continuously accepting input of image data indicative of an image and detecting a range of an image pickup object in an image based on the accepted image data, characterized by comprising: means for manipulating second image data after first image data on the basis of a range of the image pickup object detected from an image based on the first image data; and means for detecting a range of the image pickup object in an image based on the manipulated second image data.
A computer program according to the ninth aspect is a computer program for causing a computer, which comprises means for continuously accepting input of image data indicative of an image, to detect a range of an image pickup object in an image based on the accepted image data, characterized by running: a procedure of causing a computer to manipulate second image data after first image data on the basis of a range of the image pickup object detected from an image based on the first image data; and a procedure of causing a computer to detect a range of the image pickup object in an image based on the manipulated second image data.
With the first aspect, the second aspect, the seventh aspect, the eighth aspect and the ninth aspect which are, for example, applied to a system using an in-vehicle camera mounted on a vehicle for an image pickup object of the face of the driver and properly manipulate a second image on the basis of the processing result for a first image, it is possible to facilitate image processing, such as detection of a range related to an image object, for the second image. Especially, when being applied to the image pickup device shown in the seventh aspect wherein an image is preliminarily manipulated as pretreatment and outputted to the image processing device so that the image pickup object is shown in the image in a standard state, it is possible to reduce the processing load on the image processing device and to use the image pickup device in a plurality of types of image processing software in common, so that it is possible to reduce memory for recording image data. Moreover, with the eighth aspect and the ninth aspect, it is possible to realize image processing of the present invention using a conventional image pickup device.
With the third aspect wherein an image is also enlarged when, for example, the face of the driver which is the image pickup object is small, image processing such as detection of a range related to the image object is facilitated.
With the fourth aspect wherein an enlargement ratio is properly decided, it is possible to adjust the size of the image pickup object to a size suitable for image processing.
With the fifth aspect, it is possible to adjust the image pickup object to location which facilitates image processing even when the image pickup object moves in a taken image.
The sixth aspect can be applied to detection of the facial contour of the driver who drives a vehicle, the location of the eyes and nose and also the status of the face based thereon.
An image processing method, an image processing system, an image pickup device, an image processing device and a computer program according to the present invention are applied, for example, to a system using an in-vehicle camera mounted on a vehicle for an image pickup object of the face of the driver so as to: continuously take an image of the image pickup object; perform manipulation, such as enlargement of an area including the image pickup object with a reference point being the center, for a second image based on image pickup after a first image on the basis of the range of the image pickup object detected from the first image, the width of the facial contour of the driver, for example, and also a reference point such as the center of the face to be decided based on the location of the eyes and nose of the driver; and perform image processing, such as detection of the range of the image pickup object and decision of the reference point, for the manipulated second image.
With this structure, since an image pickup object in an image can be enlarged when the same is small and the location can be adjusted, beneficial effects can be guaranteed such as facilitation of image processing such as detection of the range of the image pickup object and decision of a reference point.
Especially, when being applied to an image pickup device such as an in-vehicle camera, since the processing load on an image processing device connected with the image pickup device can be reduced and the image pickup device can be used in a plurality of types of image processing software in common by preliminarily manipulating an image as pretreatment and outputting the image to the image processing device so that the image pickup object is shown in the image in a standard state, beneficial effects can be guaranteed such as reduction of memory for recording image data.
1 Image Pickup Device
2 Image Processing Device
3 Computer Program
4 Memory Product
The following description will explain the present invention in detail with reference to the drawings illustrating some embodiments thereof.
The image pickup device 1 comprises: an MPU (Micro Processor Unit) 11 for controlling the entire device; a ROM (Read Only Memory) 12 for recording data and various kinds of computer programs to be executed based on control of the MPU 11; a RAM (Random Access Memory) 13 for storing various kinds of data to be generated temporally during execution of a computer program recorded in the ROM 12; an image pickup unit 14 constituted of an image pickup element such as a CCD (Charge Coupled Device); an A/D converter 15 for converting analog image data obtained by image pickup of the image pickup unit 14 into digital data; a processing circuit 16 for performing image processing, which will be explained later, for digital image data converted by the A/D converter 15; a frame memory 17 for temporally storing image data processed by the processing circuit 16; and a communication interface 18 to be used for communication with the image processing device 2.
In the image pickup device 1, the image pickup unit 14 performs an image pickup process continuously or intermittently so as to generate 30 image data (image frames) per second, for example, on the basis of an image pickup process and output the same to the A/D converter 15, and the A/D converter 15 converts each pixel constituting image data into digital image data which is shown by a gradation such as 256 gradation (1 Byte). The image data processed by the processing circuit 16 is then stored in the frame memory 17 and the stored image data is outputted to the image processing device 2 at predetermined timing. Moreover, the image pickup device 1 accepts various kinds of data necessary for a process of the processing circuit 16 from the image processing device 2 at predetermined timing and the accepted data is recorded in a memory such as a flash memory provided in the processing circuit 16. It should be noted that various kinds of data such as a reference value necessary for image processing to be performed by the processing circuit 16 is preliminarily recorded in the memory.
The image processing device 2 comprises: a CPU (Central Processing Unit) 21 for controlling the entire device; a hard disk (which will be hereinafter referred to as an HD) 22 for recording data and various kinds of computer programs to be executed based on control of the CPU 21; a RAM 23 for storing various kinds of data to be generated temporarily during execution of a computer program recorded in the HD 22; a frame memory 24 constituted of, for example, a nonvolatile storage; a communication interface 25 to be used for communication with the image pickup device 1; and an auxiliary storage unit 26 for accessing a record medium such as a CD-ROM or a memory card.
In the image processing device 2, image data outputted from the image pickup device 1 is accepted at the communication interface 25 and a computer program recorded in the HD 22 is executed for the accepted image data, so that a variety of image processing is performed and various kinds of data based on the result of image processing is outputted from the communication interface 25 to the image pickup device 1. Image processing based on execution of a computer program recorded in the HD 22 is a variety of processes such as: detection of the facial contour of the driver, which is the image pickup object, from image data; identification of the location of the eyes and nose; and also detection of the orientation of the face based on the result thereof. A concrete example of a process is a process of detecting the location of the eyes by pattern matching. Another example of a process is a process of integrating the brightness on a vertical line of an image and comparing the integrated value with a predetermined threshold so as to detect the horizontal range of the facial contour composed of pixels having higher brightness than background. Furthermore, another example is a process of differentiating a horizontal change in the integrated value in the above process to identify the location having a large change and detecting a boundary between background and the facial contour where the brightness greatly changes. The detailed content of the processes is described in documents such as Japanese Patent Application Laid-Open Nos. 2000-163564, 2004-234494 and 2004-234367 which have been applied by the present applicant. It should be noted that the above image processing is not always limited to the processes described in Japanese Patent Application Laid-Open Nos. 2000-163564, 2004-234494 and 2004-234367, and can be selected properly according to the terms such as the use thereof, the hardware structure or cooperation with other application programs.
The following description will explain various kinds of processes to be performed in an image processing system in Embodiment 1 of the present invention.
When it is determined in the step S202 that the accepted range data is larger than the maximum value of range data (S202: YES), the image pickup device 1 performs an enlargement ratio changing process of decreasing a numerical value indicative of an enlargement ratio recorded in the memory of the processing circuit 16, under control of the MPU 11 (S203). A numerical value indicative of an
When it is determined in the step S202 that the accepted range data is smaller than or equal to the maximum value of range data (S202: NO), the image pickup device 1 compares the accepted range data with the minimum value of range data, which is preliminarily recorded in the memory of the processing circuit 16 as a reference value, under control of the MPU 11 (S204).
When it is determined in the step S204 that the accepted range data is smaller than the minimum value of range data (S204: YES), the image pickup device 1 performs an enlargement ratio changing process of increasing a numerical value indicative of an enlargement ratio recorded in the memory of the processing circuit 16, under control of the MPU 11 (S205). That is, an enlargement ratio is decided based on the size of the range indicated by the range data in the processes of the steps S202-S205. It should be noted that an enlargement ratio may be changed based on accepted range data every time range data is accepted, without comparing the range data with the maximum value and the minimum value. It should be noted that, when the image processing device 2 does not convert the coordinate system for accepted range data, a conversion process is performed before comparison of the step S202.
After an enlargement ratio changing process is performed in the step S203 or the step S205, or when it is determined in the step S204 that the accepted range data is larger than or equal to the minimum value of range data (S204: NO), the image pickup device 1 decides coordinates to be the center of enlargement on the basis of the reference point data accepted in the step S201 under control of the MPU 11 (S206) and records the decided coordinates to be the center of enlargement into the memory of the processing circuit 16 (S207). In the step S206, the reference point data is decided as coordinates to be the center of enlargement. Here, when the image processing device 2 does not convert the coordinate system for the accepted reference point data, the coordinates after the conversion of the coordinate system of the reference point data is decided as coordinates to be the center of enlargement. Moreover, for decision of coordinates to be the center of enlargement, whether the said coordinates is within the effective range of an enlarged image or not is also determined based on an enlargement ratio recorded in the memory.
Embodiment 1 of the present invention.
The image processing device 2 then accepts input of enlarged image data after the manipulation process at the communication interface 25 and performs an image recognition process shown using
Embodiment 2 is an embodiment of performing a manipulation process of image data in Embodiment 1 at the image processing device 2.
The image processing device 2 accepts the image data outputted from the image pickup device 1 at the communication interface 25 under control of the CPU 21 (S402) and stores the accepted image data in the frame memory 24 (S403).
The image processing device 2 then performs a recognition process of an image pickup object, which is image processing including detection of the facial contour of the driver that is the image pickup object from enlarged image data and identification of the location of the eyes and nose, under control of the CPU 21 (S503) and supports driving on the basis of the result of the recognition process of the image pickup object. The image processing device 2 further detects the range of the image pickup object on the basis of the result of the recognition process under control of the CPU 21 (S504) and then decides a reference point in the image pickup object (S505).
The image processing device 2 then compares the range of the image pickup object detected in the step S504 with the maximum value of the range of the image pickup object, which is preliminarily recorded in the HD 22 as a reference value, under control of the CPU 21 (S506).
When it is determined in the step S506 that the detected range of the image pickup object is larger than the recorded maximum value of the range (S506: YES), the image processing device 2 performs an enlargement ratio changing process of decreasing a numerical value indicative of an enlargement ratio, which is recorded in the HD 22 or the RAM 23; under control of the CPU 21 (S507).
When it is determined in the step S506 that the detected range of the image pickup object is smaller than or equal to the recorded maximum value of the range (S506: NO), the image processing device 2 compares the range of the image pickup object detected in the step S504 with the minimum value of the range of the image pickup object, which is preliminarily recorded in the HD 22 as a reference value, under control of the CPU 21 (S508).
When it is determined in the step S508 that the detected range of the image pickup object is smaller than the recorded minimum value of the range (S508: YES), the image processing device 2 performs an enlargement ratio changing process of increasing a numerical value indicative of an enlargement ratio, which is recorded in the HD 22 or the RAM 23, under control of the CPU 21 (S509).
After performing an enlargement ratio changing process in the step S507 or the step S509, or when it is determined in the step S508 that the detected range of the image pickup object is larger than or equal to the recorded minimum value of the range (S508: NO), the image processing device 2 decides coordinates to be the center of enlargement on the basis of the reference point decided in the step S505, under control of the CPU 21 (S510) and records the decided coordinates to be the center of enlargement into the HD 22 or the RAM 23 (S511). An enlargement ratio and coordinates to be the center of enlargement which are recorded in the HD 22 or the RAM 23 when the processes to the step S511 are executed are used for a manipulation process of enlarging the next image data.
As explained in Embodiments 1 and 2 in detail, the present invention repeatedly performs a process of: continuously taking an image of an image pickup object; manipulating a second image obtained by image pickup after a first image on the basis of an enlargement ratio and coordinates to be the center of enlargement decided based on the first image; performing a variety of image processing based on the manipulated second image; and also deciding an enlargement ratio and coordinates to be the center of enlargement of the next image.
Though a process of enlarging an image is shown in the above Embodiments 1 and 2 as a process for manipulating an image, the present invention is not limited to this and a variety of processes such as reduction, conversion of gradation or a differentiation process can be used. Moreover, though an embodiment of enlarging image data is shown as an enlargement process, an embodiment of enlarging the range of image pickup optically can be also employed.
Though an embodiment for an image pickup object of the driver of a vehicle is shown in the above Embodiments 1 and 2, the present invention is not limited to this and an embodiment for an image pickup object of a variety of human, or even a living matter other than human or a nonliving matter can be also employed.
Number | Date | Country | |
---|---|---|---|
Parent | PCT/JP2005/002467 | Feb 2005 | US |
Child | 11889990 | Aug 2007 | US |