JP2007251258A - Image recognizing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve detection precision in an object around an own vehicle and precision in the determination of a situation by coping with a backlight state. <P>SOLUTION: A specific light source detector 21 estimates whether an image is a specific light source, such as a headlight and a streetlamp, for example, from the size, shape, and a difference in luminance to the periphery of the image having luminance of at least a prescribed value in the image photographed by a camera 31. An exposure time corrector 22 corrects exposure control in the camera 31 so that optimum exposure time is used for an image from which the image of the specific light source is excluded. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image recognition apparatus that performs image recognition on an image captured by a camera, and more particularly to an image recognition apparatus that can perform image recognition with high accuracy even when a headlight or the like of an oncoming vehicle is present.

  2. Description of the Related Art In recent years, a technique has been devised that assists a driving operation by a driver by photographing the surroundings of the host vehicle and recognizing the presence of a pedestrian or vehicle by image recognition. For example, Patent Literature 1 discloses a technique for recognizing a pedestrian image from an infrared image.

  In addition, when an image taken for image recognition is in a backlit state, the exposure time becomes excessively short, and sufficient exposure of a portion to be viewed cannot be obtained, and the surrounding objects are darkened. Therefore, in Patent Document 2, the inside of an image is divided into predetermined areas, and exposure control is performed on the area to be noticed without being affected by the backlight outside the area of interest.

JP 2005-159392 A JP 2004-88545 A

  However, when recognizing whether or not a specific object exists in the image, it is not known where the target object (for example, a pedestrian) is located in the image. If there are pedestrians, they will be affected by backlighting.

  Therefore, even if the light from the specific light source is incident on the on-vehicle camera, such as when there is a headlight or streetlight of an oncoming vehicle, it is not affected by the backlight, and it is accurate regardless of where the specific object is located in the image. Recognition was an important issue.

  The present invention has been made to solve the above-described problems in the prior art and to solve the problems, and provides an image recognition device that improves the accuracy of detection of objects around the host vehicle and the accuracy of situation determination. With the goal.

  In order to solve the above-described problems and achieve the object, an image recognition apparatus according to the invention of claim 1 is an image recognition apparatus that performs image recognition of a specific object with respect to an image taken by a camera. A specific light source detecting means for detecting an image of the specific light source from the camera, and based on a detection result by the specific light source detecting means, the camera is configured to use an optimal exposure time for an image excluding the image of the specific light source. And a photographing setting control means for controlling the setting.

  According to the first aspect of the present invention, when the image of the specific light source is present in the image captured by the camera, the image recognition apparatus uses the optimum exposure time for the image excluding the image of the specific light source. Control.

  An image recognition apparatus according to a second aspect of the present invention is the image recognition apparatus according to the first aspect, wherein the specific light source is a light source that repeatedly turns on and off at a specific cycle, and the photographing setting control means The shooting setting of the camera is changed so that the shooting interval is shorter than the specific period.

  According to the second aspect of the present invention, the image recognition device controls the imaging interval to be shorter than the lighting period of the pulsed light source when the image of the pulsed light source exists in the image captured by the camera. .

  The image recognition apparatus according to a third aspect of the present invention is the image recognition apparatus according to the first aspect, wherein the specific light source is a light source that repeatedly turns on and off at a specific cycle, and the photographing setting control means includes: The shooting setting is controlled so that shooting is performed when the specific light source is turned off.

  According to the third aspect of the present invention, the image recognition apparatus controls the phase of the photographing timing so that photographing is performed while the pulsed light source is turned off when an image of the pulsed light source exists in the image photographed by the camera.

  According to a fourth aspect of the present invention, there is provided an image recognition apparatus for recognizing a specific object from an image captured by a camera, wherein the specific light source detection detects an image of a specific light source from the image. And a blocking control unit for selectively blocking incident light from the specific light source by controlling partial blocking means for partially blocking incident light on the camera based on a detection result by the specific light source detecting unit. , Provided.

  According to the fourth aspect of the present invention, the image recognition device selectively blocks incident light from the specific light source when an image of the specific light source exists in the image taken by the camera.

  According to a fifth aspect of the present invention, in the image recognition apparatus according to the fourth aspect, the blocking control means predicts the position of the image of the specific light source at the time of the next shooting, and the incident light with respect to the predicted position. Is selectively cut off.

  According to the invention of claim 5, the image recognition device predicts the position of the image of the specific light source in the next frame when the image of the specific light source exists in the image captured by the camera, and determines the incident light with respect to the predicted position. Block selectively.

  An image recognition apparatus according to a sixth aspect of the present invention is an image recognition apparatus that performs image recognition of a specific object with respect to an image captured by a camera, the image recognition result and a detection result by other detection means. And a specific light source detecting means for detecting an image of a specific light source from the image, the status determining means based on the detection means by the specific light source detecting means. In the situation determination for the vicinity of the light source, the detection result by the other detection means is preferentially used.

  According to the sixth aspect of the present invention, when the image of the specific light source is present in the image captured by the camera, the image recognition device prioritizes the detection result by the other detection means in the situation determination for the vicinity of the specific light source. use.

  An image recognition apparatus according to a seventh aspect of the present invention is the image recognition apparatus according to any one of the first to sixth aspects, wherein the specific light source detection means has a size and shape of an image having a luminance of a predetermined value or more. It is characterized in that it is estimated whether or not the image is a specific light source image using at least one of the luminance differences from the surroundings.

  According to the seventh aspect of the present invention, the image recognizing device is an image of a specific light source using the size and shape of an image having a luminance equal to or higher than a predetermined value in the image taken by the camera, and the luminance difference from the periphery. Estimate.

  According to an eighth aspect of the present invention, in the image recognition apparatus according to the seventh aspect, the specific light source is a headlamp of another vehicle, and the specific light source detection means has a luminance greater than or equal to the predetermined value. It is further characterized in that it is estimated whether or not the image is the headlamp by further using a position of the image having the image on the image.

  According to the eighth aspect of the present invention, the image recognition apparatus uses the size and shape of an image having a luminance equal to or higher than a predetermined value in the image captured by the camera, the luminance difference from the surroundings, and the position on the image, to detect the other vehicle. It is estimated whether it is a headlight image.

  According to the first aspect of the present invention, when the image of the specific light source is present in the image captured by the camera, the image recognition device is controlled to use the optimum exposure time for the image excluding the image of the specific light source. Therefore, there is an effect that it is possible to obtain an image recognition device that improves the detection accuracy of objects around the host vehicle and the accuracy of situation determination.

  According to the second aspect of the present invention, the image recognition device controls the shooting interval to be shorter than the lighting cycle of the pulsed light source when an image of the pulsed light source exists in the image captured by the camera. Therefore, there is an effect that it is possible to obtain an image recognition device that captures images while avoiding a backlight condition and improves image recognition accuracy.

  According to a third aspect of the present invention, the image recognition device controls the phase of the photographing timing so that photographing is performed while the pulsed light source is turned off when an image of the pulsed light source exists in the image photographed by the camera. Therefore, there is an effect that it is possible to obtain an image recognition apparatus that captures images while avoiding a backlight state and improves image recognition accuracy.

  According to the invention of claim 4, the image recognition device selectively blocks incident light from the specific light source when an image of the specific light source is present in the image taken by the camera. There is an effect that it is possible to obtain an image recognition device that avoids the generation of a luminance region and secures an optimum exposure amount for the region to be recognized.

  According to the invention of claim 5, the image recognition device predicts the position of the image of the specific light source in the next frame when the image of the specific light source exists in the image captured by the camera, and the incident light with respect to the predicted position. Therefore, it is possible to obtain an image recognition apparatus that reliably avoids the generation of a high luminance area in an image and secures an optimum exposure amount for the area to be recognized.

  According to the invention of claim 6, when the image of the specific light source is present in the image taken by the camera, the image recognition device prioritizes the detection result by other detection means in the situation determination for the vicinity of the specific light source. Therefore, there is an effect that it is possible to obtain an image recognition device that improves the accuracy of detection of objects around the host vehicle and the accuracy of situation determination.

  According to the seventh aspect of the present invention, is the image recognition device an image of a specific light source using the size and shape of an image having a luminance greater than or equal to a predetermined value in the image captured by the camera, and the luminance difference from the periphery? Therefore, it is possible to obtain an image recognition device that can accurately detect a backlight state and improve the accuracy of detecting an object around the host vehicle and the accuracy of situation determination.

  According to the invention of claim 8, the image recognition apparatus uses another vehicle by using the size and shape of an image having a luminance equal to or higher than a predetermined value in the image taken by the camera, the luminance difference from the periphery, and the position on the image. Therefore, it is possible to obtain an image recognition device that avoids a backlight condition caused by a headlamp of another vehicle and improves the accuracy of detection of objects around the host vehicle and the accuracy of situation determination. There is an effect that can be.

  Exemplary embodiments of an image recognition apparatus according to the present invention will be described below in detail with reference to the accompanying drawings.

  FIG. 1 is a schematic configuration diagram showing a schematic configuration of an in-vehicle image recognition apparatus 10 that is an embodiment of the present invention. As shown in the figure, the vehicle has the following configuration. The image recognition device 10 is connected to the navigation device 30, the camera 31, the pre-crash ECU 40, the display 43, and the speaker 44.

  The navigation device 30 is an in-vehicle device that sets and guides a travel route using the position of the host vehicle specified by communicating with a GPS (Global Positioning System) artificial satellite and map data 30a stored in advance. In addition, the navigation device 30 provides the image recognition device 10 with location information of the host vehicle, surrounding map information, a planned travel route, and the like.

  The camera 31 is a photographing unit that photographs the periphery of the host vehicle and inputs the photographing result to the image recognition device 10. The camera 31 includes an exposure control unit 31a and a photographing timing control unit 31b. The exposure control unit 31a performs processing for controlling the exposure time at the time of photographing based on the luminance distribution of the photographed image. Then, the photographing timing control unit 31b performs processing for controlling the timing of executing photographing and the interval (frame rate).

  The pre-crash ECU 40 receives the control of the image recognition device 10 when the image recognition device 10 predicts a collision of the host vehicle, controls the operation of the vehicle by the brake 41 and the engine control device (EFI) 42, and displays the display 43 and the speaker. This is an electronic control device that executes notification control by the control unit 44.

  The display 43 is an output means for notifying the user, that is, the vehicle occupant by display, and the speaker 44 is an output means for notifying by voice. The display 43 and the speaker 44 can be shared by various in-vehicle devices such as the navigation device 30 and an in-vehicle audio device (not shown) in addition to performing output under the control of the pre-crash ECU 40.

  The image recognition apparatus 10 includes a preprocessing unit 11, a vehicle recognition unit 16, a white line recognition unit 17, a pedestrian recognition unit 18, a collision determination unit 19, and a specific light source handling processing unit 20 therein. Here, the vehicle recognition unit 16, the white line recognition unit 17, the pedestrian recognition unit 18, the collision determination unit 19, and the specific light source countermeasure processing unit 20 are functional units realized by the microcomputer 10a.

  The preprocessing unit 11 performs processing such as filtering, edge detection, and contour extraction on the image captured by the camera 31, and then outputs the image to the vehicle recognition unit 16, the white line recognition unit 17, and the pedestrian recognition unit 18.

  The vehicle recognition unit 16 performs pattern matching on the image output from the preprocessing unit 11 to recognize the vehicle, and outputs the recognition result to the collision determination unit 19. Further, the white line recognition unit 17 recognizes a white line by performing pattern matching or the like on the image output from the preprocessing unit 11, and outputs the recognition result to the collision determination unit 19.

  The pedestrian recognition unit 18 detects an area where a pedestrian may exist from the image output by the preprocessing unit 11 as a pedestrian candidate area, and further performs pattern matching on the pedestrian candidate area and walks. The person is recognized and output to the collision determination unit 19.

  The collision determination unit 19 uses the recognition results obtained by the vehicle recognition unit 16, the white line recognition unit 17, the pedestrian recognition unit 18, and the position information output by the navigation device 30, and the collision risk between the pedestrian and other vehicles and the host vehicle. Determine. It should be noted that other sensor outputs such as radar can be further used to determine the collision risk.

  Specifically, the collision determination unit 19 determines the probability of a collision with a pedestrian or another vehicle, the collision time, the distance to the collision position, the collision angle, and the like, and pre-crash based on the determination result A brake control instruction, an EFI control instruction, an information display instruction by the display 43, a warning sound output instruction by the speaker 44, and the like are output to the ECU 40.

  The specific light source handling processing unit 20 is a processing unit that deals with a specific light source when an image taken by the camera 31 includes an image of a specific light source such as a headlight or a streetlight. Part 21 and exposure time correction part 22.

  The specific light source detection unit 21 performs image processing on the image captured by the camera 31 and performs processing for detecting an image of the specific light source. Specifically, the specific light source detection unit 21 determines whether the image has a specific light source, for example, an image of a headlight or a streetlight, based on the size, shape, and luminance difference between the image having a luminance of a predetermined value or more in the image. It is estimated whether it is.

  Furthermore, regarding the headlight of an oncoming vehicle that has a great influence on the captured image, it is estimated whether or not the image is a headlight by further using the position on the image.

  The exposure time correction unit 22 performs a process of correcting the exposure time of the camera 31 based on the detection result of the specific light source detection unit 21. Specifically, the exposure time control operation of the exposure control unit 31a is corrected so that an optimal exposure time is used for an image excluding the image of the specific light source detected by the specific light source detection unit 21.

  The control and correction of the exposure time will be further described with reference to FIGS. First, FIG. 2 is an example of an image taken by the camera 31, and each pixel is also clearly displayed for the sake of simplicity.

  In the exposure time control by the exposure control unit 31a, the total luminance value in the vertical direction is calculated for each column for this image (for example, from the leftmost column to the rightmost column in order), and the total luminance value in one column is equal to or greater than a predetermined threshold value. Record the number of rows that become If the number of rows is equal to or greater than the threshold value, the exposure time is shortened.

  For example, in FIG. 2, if the threshold value of the total luminance value is 5000 and the threshold number of columns is 5, four columns including the image of the right headlight of the oncoming vehicle and four images including the image of the left headlight of the oncoming vehicle are provided. The total luminance value of the column exceeds the threshold value. Therefore, the number of columns exceeding the threshold value is 8, and the exposure time is shortened because the number of columns exceeds the threshold value.

  When a specific light source such as a headlight is included in the exposure control target in this way, the exposure time is shortened depending on the luminance value of the specific light source, the exposure amount is insufficient in the peripheral area to be the target of image recognition, and the image recognition accuracy is extended. The accuracy of situation judgment is reduced.

  Therefore, as shown in FIG. 3, the total luminance value in the horizontal direction is calculated for each row (for example, from the top row to the bottom row in order), and the number of columns in which the total luminance value in one row is equal to or greater than a predetermined threshold is calculated. Record. From the calculation result of the total luminance value of the row and the calculation result of the total luminance value of the column, the region of the image of the specific light source having a large luminance value can be extracted.

  Then, by performing optimal exposure control on the image excluding the image of the specific light source, a sufficient exposure time is ensured for the peripheral region to be subjected to image recognition, and the image recognition accuracy and thus the situation determination Accuracy can be improved.

  Next, the processing operation of the specific light source countermeasure processing unit 20 will be described. FIG. 4 is a flowchart for explaining the processing operation of the specific light source countermeasure processing unit 20, and is a process started when an image is acquired from the camera 31.

  When an image is acquired from the camera 31, first, the specific light source detection unit 21 detects an area showing a luminance greater than or equal to a predetermined value from the image (step S101), and specifies from the size, shape, and density difference of the detected area. It is determined whether it is a light source (step S102).

  As a result, if the specific light source does not exist in the image (No at Step S103), the process is terminated as it is. If the specific light source exists (Yes at Step S103), the region corresponding to the specific light source is excluded, The exposure time is corrected so that the optimal exposure time is obtained for the remaining area (step S104), and the process is terminated.

  On the other hand, image recognition is performed by the process shown in FIG. The flowchart shown in the figure is processing started when an image is acquired from the camera 31. First, vehicle recognition by the vehicle recognition unit 16, white line recognition by the white line recognition unit 17, and pedestrian recognition by the pedestrian recognition unit 18 are performed on the image output by the preprocessing unit 11 (step S1001). The collision determination unit 19 determines the collision risk level using the recognition result of the vehicle, the white line, and the pedestrian and the position information acquired from the navigation device 30 (step S1002) (step S1003).

  As a result, if there is no possibility of collision (step S1004, No), the processing is terminated as it is, and if there is a possibility of collision (step S1004, Yes), notification to the user or vehicle control according to the degree of danger. Is executed (step S1005), and the process is terminated.

  In this way, when a specific light source is present in the image, the image is recognized by excluding the image of the specific light source and controlling the exposure time, thereby preventing exposure shortage due to the influence of the specific light source, and recognition accuracy. Can be improved.

  It should be noted that the specific light source handling process shown in FIG. 4 and the image recognition process shown in FIG. 5 may be executed independently or configured to be executed continuously. Moreover, it is good also as a different subroutine called by the main routine inside the microcomputer 10a.

  By the way, the coping method for the specific light source is not limited to the control of the exposure time, and other methods can be used.

  FIG. 6 is a schematic configuration diagram showing a configuration in a case where incident light from a specific light source is selectively blocked when there is a specific light source.

  In the figure, the camera 31 and the liquid crystal panel 32 constitute a camera unit 45. The liquid crystal panel 32 is a partial blocking unit that partially blocks incident light to the camera 32. Further, the specific light source handling processing unit 20a selectively blocks incident light from the specific light source by controlling the liquid crystal panel 32 based on the detection result of the specific light source detection unit 21.

  Here, the position of the image of the specific light source in the image changes depending on the travel of the host vehicle or the movement of the specific light source itself. Therefore, the liquid crystal panel control unit 23 predicts the position of the image of the specific light source at the next photographing timing, and controls the liquid crystal panel 32 so as to selectively block incident light at the predicted position.

  Since other configurations and operations are the same as those shown in FIG. 1, the same components are denoted by the same reference numerals and description thereof is omitted.

  Next, the processing operation of the specific light source countermeasure processing unit 20a shown in FIG. 6 will be described. FIG. 7 is a flowchart for explaining the processing operation of the specific light source handling processing unit 20a, and is a process started when an image is acquired from the camera 31.

  When an image is acquired from the camera 31, the specific light source detection unit 21 first detects an area showing a luminance of a predetermined value or more from the image (step S <b> 201), and specifies from the size, shape, and density difference with the surrounding area of the detected area. It is determined whether it is a light source (step S202).

  As a result, if the specific light source does not exist in the image (No at Step S203), the process is terminated, and if the specific light source exists (Step S203, Yes), the position of the specific light source in the next frame is predicted. (Step S204). Then, the area where the specific light source is located in the next frame is masked by the liquid crystal panel (step S205), and the process is terminated.

  In this way, when a specific light source is present in the image, by selectively blocking the incident light from the specific light source, an excessively bright image is excluded from the image, and the exposure amount due to the influence of the specific light source Shortage can be prevented.

  Next, a configuration for avoiding the shortage of exposure amount by controlling the photographing timing when the specific light source is a pulse light source will be described with reference to FIG.

  When street lamps or headlights are constituted by LEDs, they may be pulsed light sources that repeatedly turn on and off at a specific cycle. In this case, by taking a picture while the specific light source is turned off, it is possible to avoid an insufficient exposure amount due to the specific light source.

  Specifically, in FIG. 8, the specific light source handling processing unit 20 b acquires the detection result by the optical sensor unit 34 and includes a specific light source detection unit 21, a frame rate setting unit 24, and a phase control unit 25 inside. Since other configurations and operations are the same as those shown in FIG. 1, the same components are denoted by the same reference numerals and description thereof is omitted.

  The optical sensor unit 34 is configured by a photodiode or the like, and measures the amount of light around the host vehicle. Here, the optical sensor unit 34 may measure the light amount of the entire photographing range of the camera 31 or measure the light amount for each direction within the photographing range of the camera 31, that is, for each position in the image. It may be.

  When the specific light source is a pulse light source, the pulse period can be acquired from the change in the light amount measured by the optical sensor unit 34. Therefore, when the specific light source detection unit 21 detects the specific light source, the frame rate setting unit 24 sets the shooting interval of the camera 31, that is, the frame rate, based on the pulse period of the specific light source, and the phase control unit 25 Control the phase of.

  The control of the photographing timing will be described with reference to FIG. As shown in the figure, when the blinking of the headlight of the oncoming vehicle is in the cycle f1, the headlight of the oncoming vehicle is turned off by setting the frame rate to a cycle f2 shorter than f1 by the frame rate setting unit 24. Shooting can be performed at the timing when

  Also, even when the frame rate cycle is the same as the headlight lighting cycle f1, shooting is performed at the timing when the headlight is turned off by controlling the phase of the shooting so that it is in the opposite phase of the headlight lighting. Can be performed.

  Next, the processing operation of the specific light source countermeasure processing unit 20b shown in FIG. 8 will be described. FIG. 10 is a flowchart for explaining the processing operation of the specific light source handling processing unit 20 b, and is a process that is started when an image is acquired from the camera 31.

  When an image is acquired from the camera 31, first, the specific light source detection unit 21 detects an area showing a luminance of a predetermined value or more from the image (step S301), and specifies from the size, shape, and density difference of the detected area. It is determined whether it is a light source (step S302).

  As a result, if the specific light source does not exist in the image (No at Step S303), the process is terminated as it is, and if the specific light source exists (Step S303, Yes), the pulse cycle of the specific light source from the optical sensor unit 34. Is acquired (step S304). Then, the imaging timing is controlled based on the pulse period (step S305), and the process is terminated.

  In this way, when a pulsed light source is present in the image, an image with an excessively high brightness is excluded from the image by acquiring the period of the pulsed light source and shooting at the timing when the light source is turned off. Insufficient exposure due to the influence of the can be prevented.

  In FIG. 8, the configuration in which the light amount is measured by the optical sensor unit 34 has been described as an example. However, when the configuration is simplified, a change in the light amount is calculated by image processing on an image captured by the camera 31. Thus, the optical sensor unit 34 can be omitted.

  Next, in the case where a specific light source is present, a configuration in which priority is given to the detection result by another detection means in the situation determination for the vicinity of the specific light source will be described with reference to FIG.

  In the configuration shown in the figure, the collision determination unit 19 uses the detection results obtained by the radar 33 in addition to the recognition results obtained by the vehicle recognition unit 16, the white line recognition unit 17 and the pedestrian recognition unit 18 and the position information output from the navigation device 30. The collision risk between the pedestrian or other vehicle and the host vehicle is determined.

  Here, the radar 33 is a detection means for detecting an object around the own vehicle and measuring a distance to the object by using reflection of radio waves. In addition, although the case where a radar is used is described here as an example, any other detection means can be used.

  The specific light source handling processing unit 20c includes a specific light source detection unit 21 and a determination priority setting unit 26 therein. When the specific light source detection unit 21 detects the specific light source, the determination priority setting unit 26 instructs the collision determination unit 19a to preferentially use the detection result of the radar 33 in the situation determination around the specific light source.

  Next, the processing operation of the specific light source countermeasure processing unit 20c shown in FIG. 11 will be described. FIG. 12 is a flowchart for explaining the processing operation of the specific light source handling processing unit 20 c, and is a process that is started when an image is acquired from the camera 31.

  When an image is acquired from the camera 31, the specific light source detection unit 21 first detects an area showing a luminance of a predetermined value or more from the image (step S 401), and specifies the size, shape, and density difference from the surrounding area of the detected area. It is determined whether it is a light source (step S402).

  As a result, if the specific light source does not exist in the image (No at Step S403), the process is ended as it is, and if the specific light source exists (Yes at Step S403), another region is present for the region where the specific light source exists. The collision determination unit 19a is instructed to give priority to the sensor output (step S404), and the process ends.

  Thus, when a specific light source is present in the image, the accuracy of the situation determination can be improved by prioritizing the detection result by other detection means in the surrounding situation determination.

  As described above, the image recognition apparatus 10 according to the present embodiment detects a specific light source such as a headlight or a streetlight from an image captured by the camera 31, and corrects an exposure time when a specific light source exists. Even if a high-intensity light source is present in the image by selectively blocking incident light, controlling shooting timing, and prioritizing other detection means in situation determination, Can be detected with high accuracy.

  As is apparent from the examples, the present invention is particularly effective for the oncoming vehicle in which the headlights are turned on, that is, image recognition at night, and more preferably, the near-infrared ray is irradiated forward from the own vehicle, A high effect is exhibited when recognizing an image from an infrared image that is the reflected light, but it can also be applied to, for example, image recognition in the visible light region or daytime image recognition.

  Further, although the present embodiment has been described as on-vehicle image recognition, the present invention is not limited to this, and can be applied to, for example, a traffic flow monitoring device or a security monitoring device.

  As described above, the image recognition apparatus according to the present invention is useful for image recognition in a vehicle, and is particularly suitable for situation determination in a backlight state.

1 is a schematic configuration diagram illustrating a schematic configuration of an image recognition apparatus according to an embodiment of the present invention. It is explanatory drawing explaining control of exposure time. It is explanatory drawing explaining correction | amendment of exposure time. It is a flowchart explaining the process operation | movement of the specific light source countermeasure process part 20 shown in FIG. It is a flowchart explaining an image recognition process. It is a schematic block diagram which shows the structure in the case of selectively interrupting incident light. It is a flowchart explaining the processing operation of the specific light source countermeasure process part 20a shown in FIG. It is a schematic block diagram which shows the structure in the case of imaging | photography timing control. It is explanatory drawing explaining the lighting period and imaging | photography timing control of a specific light source. It is a flowchart explaining the process operation | movement of the specific light source countermeasure process part 20b shown in FIG. It is a schematic block diagram which shows a structure in the case of giving priority to the detection result by another detection means. It is a flowchart explaining the processing operation of the specific light source countermeasure processing unit 20c shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image recognition apparatus 10a Microcomputer 16 Vehicle recognition part 17 White line recognition part 18 Pedestrian recognition part 19, 19a Collision determination part 20, 20a, 20b, 20c Specific light source countermeasure processing part 21 Specific light source detection part 22 Exposure time correction part 23 Liquid crystal panel Control unit 24 Frame rate setting unit 25 Phase control unit 26 Determination priority setting unit 30 Navigation device 30a Map data 31 Camera 31a Exposure control unit 31b Shooting timing control unit 32 Liquid crystal panel 33 Radar 34 Optical sensor unit 40 Pre-crash ECU
41 Brake 42 EFI
43 Display 44 Speaker 45 Camera unit

Claims (8)

An image recognition apparatus that performs image recognition of a specific object for an image captured by a camera,
Specific light source detection means for detecting an image of the specific light source from the image;
Shooting setting control means for controlling shooting settings of the camera so as to use an optimal exposure time for an image excluding the image of the specific light source based on a detection result by the specific light source detection means;
An image recognition apparatus comprising:   The specific light source is a light source that repeatedly turns on and off at a specific cycle, and the shooting setting control means changes the shooting setting of the camera so that the shooting interval is shorter than the specific cycle. The image recognition apparatus according to claim 1.   The specific light source is a light source that repeatedly turns on and off at a specific cycle, and the shooting setting control means controls shooting settings so that shooting is performed at a timing when the specific light source is turned off. The image recognition apparatus according to claim 1. An image recognition apparatus that performs image recognition of a specific object for an image captured by a camera,
Specific light source detection means for detecting an image of the specific light source from the image;
Based on a detection result by the specific light source detection means, a partial blocking means for partially blocking incident light to the camera, and a blocking control means for selectively blocking incident light from the specific light source;
An image recognition apparatus comprising:   The image recognition apparatus according to claim 4, wherein the blocking control unit predicts the position of the image of the specific light source at the time of next shooting and selectively blocks incident light at the predicted position. An image recognition apparatus that performs image recognition of a specific object for an image captured by a camera,
Situation judging means for judging the surrounding situation using the image recognition result and the detection result by other detecting means,
Specific light source detection means for detecting an image of the specific light source from the image;
And the situation determination means preferentially uses the detection result by the other detection means in the situation determination for the vicinity of the specific light source based on the detection means by the specific light source detection means. apparatus.   The specific light source detection means estimates whether or not the image is a specific light source image using at least one of a size, a shape, and a luminance difference between the image having a luminance equal to or higher than a predetermined value. The image recognition apparatus according to any one of claims 1 to 6.   The specific light source is a headlamp of another vehicle, and the specific light source detection means further uses a position on the image of an image having a luminance equal to or higher than the predetermined value to determine whether the image is the headlamp. The image recognition apparatus according to claim 7, wherein the image is estimated.

Images