JP4795930B2 - Image processing apparatus and image processing method - Google Patents

Description

  The present invention relates to an image processing apparatus and an image processing method for processing captured image data of a host vehicle imaged by a fixed focus type in-vehicle camera in which an imaging lens is set at a predetermined position.

  In recent years, in order to support safe driving of a vehicle, a captured image in front of the vehicle captured by a digital camera (hereinafter referred to as an in-vehicle camera) mounted on the host vehicle is processed to extract an edge (contour) in the image, A vehicle travel support system that recognizes a subject such as a preceding vehicle, an obstacle, or a road surface display based on the shape of the edge and controls the throttle and brake of the host vehicle has been proposed. Such a vehicle travel support system is required to perform image recognition with high accuracy in a range from a close distance in front of the vehicle to a distance of about several tens of meters. The accuracy of image recognition is greatly affected by the sharpness of the edge extracted from the captured image, that is, the sharpness of the captured image.

  As means for capturing a clear image of an object, an autofocus mechanism that automatically adjusts the focal length with respect to the distance of the object by driving an imaging lens is known. Patent Document 1 includes such a mechanism. An example of a digital camera is described. However, the provision of such an autofocus mechanism complicates the structure of the in-vehicle camera, which increases the cost. In addition, a captured image of the in-vehicle camera is captured for the purpose of recognizing a preceding vehicle, an obstacle, or the like by the image processing device, not for display output to the driver. Therefore, it is desirable to capture as many subjects as possible with a sharpness that does not hinder image recognition, rather than capturing a specific subject in focus.

Therefore, in the vehicle driving support system, the imaging lens that forms the subject image is set at a predetermined position, and the focus is fixed so that the subject within a certain distance range can be imaged with a certain degree of sharpness. A fixed focus type in-vehicle camera is used.
JP 2003-107333 A

  In the above-described fixed focus type in-vehicle camera, there is a case where the setting position of the imaging lens is slightly shifted due to vibration or impact caused by traveling of the vehicle. Then, the initially set focus is deviated and the focus is biased to a short distance or a long distance, and the image of the out-of-focus portion becomes extremely unclear. Then, the subject is not accurately recognized from such a blurred image, and as a result, it interferes with support for safe driving of the vehicle. Therefore, when such defocusing occurs, it is desirable for the driver to move the vehicle to the maintenance shop and to check and maintain the in-vehicle camera.

  However, since the captured image is processed by the image processing apparatus and is not normally displayed and output, the driver has no opportunity to see it, and the driver does not notice the deterioration of the captured image due to defocusing during driving. Therefore, in order to determine the defocus, a chart on which a graphic pattern for inspection is drawn is used to capture an inspection chart installed at a predetermined distance from the in-vehicle camera, and feature data of the captured image Therefore, it is necessary to determine whether the imaging state has deteriorated by comparing the characteristic data of the captured image of the same chart obtained at the initially set focus. However, it is also difficult to expect the driver to carry out such complicated operations such as carrying such a chart, placing it at an accurate position, and taking an image with an in-vehicle camera.

  SUMMARY OF THE INVENTION An object of the present invention is to provide an image processing apparatus and an image processing method capable of determining the defocus of a fixed focus type in-vehicle camera without requiring a complicated operation.

  In order to achieve the above object, according to the first aspect of the present invention, the captured image data of the own vehicle imaged by a fixed focus type in-vehicle camera in which the imaging lens is set at a predetermined position is processed. In the image processing apparatus, a storage unit that stores first feature data of a part of a captured image of the host vehicle captured by the in-vehicle camera when the imaging lens is at the set position, and the first feature An image processing unit that extracts second feature data of a part of the captured image of the host vehicle that is captured by the in-vehicle camera after the data is stored. The image processing unit compares the first and second feature data, and determines a defocus of the in-vehicle camera based on the comparison result.

  Further, according to the second aspect of the present invention, in the image processing apparatus for processing captured image data around the own vehicle imaged by a fixed focus type in-vehicle camera in which the imaging lens is set at a predetermined position, A storage unit that stores first feature data of a captured image of a subject other than the subject vehicle captured by the in-vehicle camera when the imaging lens is at the set position for each of a plurality of distances between the subject vehicle and the subject. And an image processing unit that extracts second feature data of a captured image of the subject imaged by the in-vehicle camera after the first feature data is stored. The image processing unit compares the first feature data corresponding to the distance data between the subject vehicle detected by the subject distance detection unit and the subject, and the second feature data, and compares the comparison data. Based on the result, the defocus of the in-vehicle camera is determined.

  Further, according to the third aspect of the present invention, in the image processing apparatus for processing captured image data around the own vehicle imaged by a fixed focus type in-vehicle camera in which the imaging lens is set at a predetermined position, A storage unit that stores first feature data of a captured image of a subject other than the vehicle that is captured by the in-vehicle camera from a predetermined imaging point when the imaging lens is at the set position; and the first feature data When the host vehicle position detection unit detects that the host vehicle is located at the imaging point after the information is stored, second feature data of the captured image of the subject captured by the in-vehicle camera is extracted. And an image processing unit. The image processing unit compares the first and second feature data, and determines a defocus of the in-vehicle camera based on the comparison result.

  According to the first aspect, when the imaging lens of the in-vehicle camera is at the initial set position, a part of the own vehicle such as an emblem on the hood with a fixed distance from the in-vehicle camera is used instead of the inspection chart. An image is taken and the feature data of the subject image is stored as reference data. After that, feature data of a part of the captured image of the host vehicle is extracted from the captured image data and compared with reference data. Therefore, it is possible to determine the defocus of the in-vehicle camera without the driver performing an operation of setting the inspection chart at a predetermined position and taking an image.

  According to the second aspect, when the imaging lens is at the initial set position, a preceding vehicle other than the own vehicle is imaged as a subject instead of the inspection chart, and a plurality of distances between the own vehicle and the subject are captured. The feature data of the subject image is extracted, and the feature data is stored as reference data. Then, feature data of the subject image is extracted from the captured image data captured and compared with reference data corresponding to the distance between the host vehicle and the subject. Therefore, even if it is difficult to set a part of the host vehicle as a subject due to the shape of the host vehicle or the mounting position of the in-vehicle camera, the driver does not perform operations such as installation of an inspection chart or imaging. It is possible to determine the defocusing of the lens.

  According to the third aspect, when the imaging lens is at the initial set position, a subject such as a known inspection chart is imaged from a predetermined imaging point, and feature data of the subject image is extracted, Feature data is stored as reference data. Thereafter, when the host vehicle is located at a predetermined imaging point, the subject is imaged from that point, and feature data of the subject image is extracted and compared with the reference data. Therefore, by moving the vehicle to a predetermined imaging point, it is possible to determine the defocus of the in-vehicle camera without the driver performing an operation of setting the inspection chart at a predetermined position and imaging.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the technical scope of the present invention is not limited to these embodiments, but extends to the matters described in the claims and equivalents thereof.

  FIG. 1 is a diagram illustrating an example in which the in-vehicle camera 2 and the image processing device 3 according to this embodiment are mounted on a vehicle. FIG. 2 is a diagram illustrating the configuration of the in-vehicle camera 2 and the image processing device 3 in the present embodiment. The in-vehicle camera 2 and the image processing device 3 integrally form an image sensor 4, and the image sensor 4 is installed on the upper part of the windshield of the vehicle 1 as shown in FIG. 1. From this position, the in-vehicle camera 2 images the preceding vehicle, various obstacles, road surface lane display, and the like in the imaging range 2a in front of the vehicle through the windshield. And based on the imaged image data by the vehicle-mounted camera 2, the image processing apparatus 3 performs the edge extraction process and image recognition process which are mentioned later.

  The in-vehicle camera 2 includes an imaging lens 21 that forms a subject image, and an imaging unit 22 that is configured by a CMOS image sensor or the like and that captures a subject image input from the imaging lens 21 and outputs captured image data. The imaging lens 21 is set at a predetermined position. Therefore, the focal point of the in-vehicle camera 2 defined by the distance between the imaging lens 21 and the imaging unit 22 is fixed at a predetermined distance and position within the imaging range 2a. Is set. That is, the in-vehicle camera 2 is configured as a fixed focus type camera.

  The image processing unit 32 of the image processing device 3 processes the image data of the captured image output from the imaging unit 22 and extracts an edge from the captured image. Here, “edge” refers to a light / dark boundary line of a captured image, and a contour line of a subject or a line that forms a shape or a pattern of a region inside the contour and the contour corresponds to the edge. Therefore, the image processing unit 32 obtains a change point of light and darkness of the image by obtaining a difference in density gradation from the adjacent pixel of each pixel, and extracts a change point of light and dark difference of a certain level or more as an edge. Then, the image processing unit 32 performs image recognition of the subject by collating the extracted edge shape with the edge shape pattern data stored in the storage unit 33. The image processing unit 32 is configured by a microcomputer that operates with an image processing program stored in a built-in ROM. Alternatively, the image processing unit 32 includes an LSI for edge extraction processing that performs edge extraction processing and a microcomputer that performs image recognition. The storage unit 33 includes a rewritable nonvolatile storage device.

  Image recognition result data by the image processing unit 32 is input to a control ECU (Electronic Control Unit) 6 of the vehicle 1. The control ECU 6 is composed of a microcomputer, and the image recognition result data input from the image processing unit 32 of the image sensor 4, the state data of the object of the host vehicle input from the vehicle-mounted radar 5 described later, or various sensors not shown. A control signal is output to the ECU that controls the throttle, brake, etc. of the vehicle 1 based on the running state data of the vehicle 1 detected by the above. As a result, traveling of the vehicle 1 is controlled, and the vehicle 1 follows the preceding vehicle and avoids obstacles.

  In the present embodiment, as described above, the imaging lens 21 is set at a predetermined position. However, there is a case where the setting position of the imaging lens 21 is shifted due to vibration or impact caused by traveling of the vehicle. Then, the initial focus is shifted, and the focus is shifted to a short distance or a long distance. Then, when the focal point is biased to a short distance, the sharpness of a long-distance subject image is lowered, and when the focal point is biased to a long distance, the sharpness of a short-distance subject image is lowered. Then, the edge extracted from the unclear image also becomes unclear.

  Here, the state of the edge extracted from the captured image will be described with reference to FIG. FIG. 3A schematically illustrates a captured image in front of the vehicle captured by the in-vehicle camera 2. In FIG. 3A, the contour of the preceding vehicle 100 that is the subject is expressed by a light / dark difference between the color of the subject image and an image other than the subject. Further, the contour of the road surface 110 is also expressed by a light / dark difference between the road surface 110 and other images. Therefore, when edge extraction is performed from the captured image of FIG. 3A by the image processing unit 32, when the focus of the in-vehicle camera 2 is in an initial setting, a change in brightness difference more than a certain level as shown in FIG. Points are extracted as sharp edges.

  In contrast, when the in-vehicle camera 2 is out of focus, the difference in brightness of the captured image is blurred, so that the edges are thick and unclear as shown in FIG. That is, the amount of edge extraction increases. Thus, the defocus of the in-vehicle camera 2 is recognized as an increase in the amount of edge extraction. Therefore, in general, in order to determine the defocus, a chart placed at a predetermined distance from the in-vehicle camera 2 is used as a subject using a chart on which an inspection graphic pattern is drawn. It is necessary to compare the extracted edge extraction amount with the edge extraction amount of the subject image captured at the initially set focus. Then, when an edge having an amount exceeding the edge extraction amount at the initial setting is extracted, it is determined that the focus is shifted.

  However, it is difficult to expect the driver to carry out such a complicated operation regularly and spontaneously, such as carrying the inspection chart, placing it at an accurate position, and taking an image with the in-vehicle camera 2. Therefore, in the present embodiment, the defocus of the in-vehicle camera 2 is determined as follows.

  FIG. 4 is a diagram illustrating a first example of the present embodiment. FIG. 5 is a diagram for explaining the operation procedure of the image processing apparatus 3 and the like in the first embodiment. In the first embodiment, as shown in FIG. 4A, the in-vehicle camera 2 is installed so that the emblem E1 attached to the hood of the vehicle 1 is captured as a part of the image. Here, both the emblem E1 and the in-vehicle camera 2 are fixed to the vehicle 1, and the distance between them is always constant. Therefore, although the image and background image of the preceding vehicle 100 change from moment to moment, the image of the emblem E1 is always constant as long as the initial focus of the in-vehicle camera 2 is not deviated. That is, a sharp edge is extracted from the captured image of the emblem E1 as shown in FIG. However, if the in-vehicle camera 2 is out of focus, an unclear thick edge is extracted from the photographed image of the emblem E1 as shown in FIG.

  Therefore, the edge extraction amount of the subject image is set as “feature data”, an edge is extracted from the captured image of the emblem E1 captured at the initial focus, and the edge extraction amount is stored in the storage unit 33 as a reference value. (S51 in FIG. 5). Thereafter, the emblem E1 is periodically imaged by the in-vehicle camera 2 during traveling of the vehicle (S52 in FIG. 5). Then, the image processing unit 32 extracts an edge from the captured image (S53 in FIG. 5), and compares the edge extraction amount extracted from the captured image of the emblem E1 with the edge extraction amount stored in the storage unit 33. (S54 in FIG. 5). If the extracted edge amount is larger (YES in S55 of FIG. 5), the image processing unit 32 determines the defocus of the in-vehicle camera 2. Then, the image processing unit 32 can output a warning display to the driver by the display unit 71 via the navigation device 7 (S56 in FIG. 5), and can prompt the driver to check the in-vehicle camera 2.

  In this way, the driver can know the defocus of the in-vehicle camera 2 in a timely manner without carrying out complicated operations such as carrying the inspection chart, placing it at an accurate position, and performing imaging.

  FIG. 6 is a diagram for explaining a second example of the present embodiment. In the second embodiment, the in-vehicle camera 2 is installed so that the display unit 71 of the navigation device 7 installed on the instrument panel 71a in the passenger compartment of the vehicle 1 is always imaged. Then, the image processing unit 32 causes the display unit 71 to display a graphic pattern for inspection via the navigation device 7. Such graphic pattern data is stored in the storage unit 33 in advance. And the vehicle-mounted camera 2 images the display part 71 which displays a figure pattern. That is, the display unit 71 that displays the graphic pattern corresponds to a part of the host vehicle that is the subject.

  Again, both the display unit 71 and the in-vehicle camera 2 are fixed to the vehicle 1, and the relative distance is always constant. Therefore, although the image and background image of the preceding vehicle 100 change every moment, the captured image of the display unit 71 displaying the graphic pattern is always constant as long as the initial focus of the in-vehicle camera 2 is not deviated.

  By utilizing this, an image of a graphic pattern displayed on the display unit 71 is picked up when the in-vehicle camera 2 has a focus set in advance, and the image processing unit 32 extracts the edge of the graphic pattern. Then, the edge extraction amount is stored in the storage unit 33. Then, the image processing unit 32 extracts the edge of the graphic pattern image displayed on the display unit 71 from the subsequent captured image, and compares the edge extraction amount with the edge extraction amount stored in the storage unit 33. And based on the comparison result, the focus shift of the vehicle-mounted camera 2 can be determined.

  Alternatively, the image processing unit 32 may be configured to calculate the resolution of the in-vehicle camera 2 from the extraction result after extracting the edge of the graphic pattern image. That is, the resolution corresponds to the feature data. In this case, since a known graphic pattern is used, reference data indicating the resolution when the in-vehicle camera 2 captures the graphic pattern of the display unit 71 with the initial focus is stored in the storage unit 33 in advance, and edge extraction is performed. The resolution obtained from the result of is compared with this. By doing so, defocus is determined when the calculated resolution is lower than the resolution indicated by the reference data. In this way, it is possible to determine the defocus of the in-vehicle camera 2 without requiring a complicated operation by the driver.

  FIG. 7 is a diagram for explaining a third example of the present embodiment. FIG. 8 is a diagram for explaining the operation procedure of the image processing apparatus 3 and the like in the third embodiment. In the third embodiment, as shown in FIG. 7A, the in-vehicle camera 2 images the preceding vehicle 100 or the like as a subject.

  First, a preceding vehicle is imaged when the in-vehicle camera 2 is initially set in focus, and a rough reference value of the amount of edge extraction of the preceding vehicle extracted from the captured image is determined for each distance between the host vehicle and the preceding vehicle. The reference value table 35 is stored in the storage unit 33 (S81 in FIG. 8). Further, the reference value table 35 may be configured such that the brightness of the captured image is used as a parameter and a reference value corresponding to a combination of brightness and distance can be obtained.

  After the reference value table 35 is stored, the vehicle-mounted camera 2 images the preceding vehicle 100 while the vehicle is traveling (S82 in FIG. 8). Then, the image processing unit 32 extracts the edge extraction amount of the subject image as feature data (S83 in FIG. 8).

  Here, the “subject distance detection unit” that detects the distance between the host vehicle 1 and the preceding vehicle 100 detects the distance between the host vehicle 1 and the preceding vehicle 100 (S84 in FIG. 8). 1, 2 and 7B, the subject distance detection unit is mounted in the vicinity of the front grille on the front surface of the vehicle 1, emits a radar beam in a certain angular direction 5a in front of the vehicle 1, The vehicle-mounted radar 5 detects the relative distance D1 of the vehicle 100 in front of the vehicle 1 and the angular direction by receiving the reflected beam. From the in-vehicle radar 5, the detected distance data of the distance D1 to the preceding vehicle 100 is input to the image processing unit 32.

  Based on the distance data of the preceding vehicle 100 input from the in-vehicle radar 5, the image processing unit 32 refers to the reference value table 35 using the distance as a parameter and extracts a reference value corresponding to the distance (S85 in FIG. 8). ). Then, the image processing unit 32 compares the extraction amount of the edge of the subject extracted from the captured image with the reference value extracted from the reference value table (S86 in FIG. 8). If the extracted edge extraction amount exceeds the reference value (YES in S87 in FIG. 8), it is determined that the in-vehicle camera 2 is out of focus, and the determination result is output (S88 in FIG. 8).

  At this time, since the subject is moving, accurate determination may be difficult only with a captured image obtained by one imaging. Therefore, edge extraction is performed from an image captured continuously for a predetermined time, and it is determined that the focus is shifted when an edge amount exceeding the reference value is extracted within a predetermined time. By doing so, a more accurate determination can be made.

  In addition, if the defocus is large, the entire captured image becomes unclear, and the edge extraction amount as the entire image may be reduced. For example, when the focus deviation is large with respect to the edge extraction amount of the vehicle 100 at the initially set focus shown in FIG. 9A, the edge extraction amount may be extremely small as shown in FIG. 9B. . However, if it is determined that the focus shift is performed only when the edge extraction amount exceeds the reference value, the focus shift is not determined when the edge extraction amount is small. Therefore, the in-vehicle radar 5 continuously observes the angle direction of the subject, and if a certain amount of edge extraction amount is not extracted at a position in the captured image corresponding to the detected angle direction, it is determined as defocusing. To do. By doing so, defocus can be determined even when the amount of edge extraction is small due to large defocus.

  As described above, according to the third embodiment, even if it is difficult to image a part of the host vehicle as a subject depending on the shape of the vehicle and the installation position of the in-vehicle camera 2, an object other than the host vehicle is detected. The focus shift of the in-vehicle camera 2 can be determined as a subject.

  FIG. 10 is a diagram for explaining a fourth example of the present embodiment. FIG. 11 is a diagram for explaining the operation procedure of the image processing apparatus 3 and the like in the fourth embodiment. In the fourth embodiment, as shown in FIG. 10, the in-vehicle camera 2 images an inspection chart 120 installed in a maintenance shop or the like as an object other than the own vehicle as a subject. When the host vehicle 1 enters the maintenance shop, the chart 120 is picked up from a predetermined pick-up point. Therefore, the picked-up image of the graphic pattern is always constant as long as the initial focus of the in-vehicle camera 2 is not deviated. .

  Therefore, first, the chart 120 is imaged from a predetermined imaging point when the in-vehicle camera 2 is initially set in advance, and the image processing unit 32 performs edge extraction from the captured image of the graphic pattern, and the amount of edge extraction is stored in the storage unit 33 (S111 in FIG. 11).

  Then, the position of the own vehicle 1 is detected by the “own vehicle position detection unit” (S112 in FIG. 11). As shown in FIG. 2, the host vehicle position detection unit includes a navigation device 7 mounted on the vehicle 1. The navigation device 7 receives the GPS signal by the communication hand unit 72, and detects the point where the host vehicle is located from the GPS signal and the map information stored in the built-in hard disk. Further, the navigation device 7 stores in advance the location of a predetermined maintenance shop where the inspection chart is installed. Therefore, the navigation device 7 detects that the host vehicle is located in the maintenance shop and notifies the image processing unit 32 of the image processing device 3.

  By doing so, the image processing unit 32 can detect that the host vehicle is located at a predetermined imaging point and the in-vehicle camera 2 and the chart 120 are fixed at a predetermined relative position (FIG. 11). S118 YES). And the chart 120 for a test | inspection is imaged with the vehicle-mounted camera 2 (S113 of FIG. 11). Then, the image processing unit 32 extracts the edge of the graphic pattern image from the captured image of the chart 120 (S114 in FIG. 11), and compares the edge extraction amount with the reference value stored in the storage unit 33 (FIG. 11 S115), the focus shift of the in-vehicle camera 2 is determined (S116 in FIG. 11), and the focus shift determination result is output (S117 in FIG. 11).

  In this case as well, as in the second embodiment, since a known figure pattern is used, the resolution of the in-vehicle camera 2 at the initially set focus is stored in advance in the storage unit 33 and obtained from the result of edge extraction. It can also be configured to compare the resolution with this.

  In this way, when the host vehicle has entered the maintenance shop and is located at a predetermined imaging point, the in-vehicle camera 2 and the maintenance shop chart 120 are fixed at a predetermined relative position. Therefore, there is no need to perform complicated operations such as installing an inspection chart by the driver himself. Then, the image processing unit 32 can determine that the captured image of the in-vehicle camera 2 has deteriorated based on the captured image of the chart 120, and can determine the defocus.

  FIG. 12 is a diagram for explaining a fifth example of the present embodiment. FIG. 12 shows a configuration in which a monitoring sensor 23 configured by a laser distance sensor that measures the distance between the imaging lens 21 and the imaging unit 22 is added to the in-vehicle camera 2 of FIG. This monitoring sensor 23 corresponds to a “lens distance measuring unit”. Then, the measured distance data from the monitoring sensor 23 is input to the image processing unit 32. Then, the image processing unit 32 compares the measured distance with the distance data when the imaging lens 21 stored in advance in the storage unit 33 is at the initial position, and the imaging lens 21 and the imaging unit 22 are compared. It is possible to detect that the distance between the two has changed, that is, the focus has shifted.

  As described above, the image processing unit 32 can directly detect the defocus, and the display unit 71 of the navigation device 7 can notify the driver of the defocus.

  As a modification of the fifth embodiment, as shown in FIG. 12, a drive mechanism 24 for driving the imaging lens 21 in the focal direction is further provided, and when the defocus is extracted, the drive mechanism 24 performs image processing. The imaging lens 21 may be driven to the initial set position in accordance with a control signal from the unit 32.

  As a sixth example of this embodiment, as shown in FIG. 2, the image processing device 3 communicates captured image data to a remote control center 8 via a communication unit 72 of the navigation device 7. By doing so, the operator can confirm the captured image with the naked eye in the control center 8, and can determine the sharpness of the image. Then, by receiving the determination result from the control center 8, the image processing apparatus 3 can output a notice of defocus. As described above, by using the judgment based on the visual sense of the operator, it is possible to detect the deterioration of the imaging state of the captured image caused by the defocus more directly and reduce the processing load of the image processing unit 32.

  In the above description, the edge extraction amount is described as the imaging state, but the change amount of the edge position in a certain time may be used as the feature data. In that case, the reference value of the change amount is stored as reference data in the storage unit 33, and if the change amount of the edge position is larger than the reference value, it can be determined that the captured image has deteriorated. Can be determined.

In each of the above-described embodiments, the case where the reference data is stored in the storage unit 33 has been described. However, the table storing the rough reference values in the third embodiment, the reference values of the edge amounts of known graphic patterns in the second and fourth embodiments, etc. are stored in advance in the built-in ROM of the image processing unit 32. It is good also as a structure written in.
As described above, according to the present invention, it is possible to determine the defocus of a fixed focus type in-vehicle camera without requiring a complicated operation.

It is a figure explaining the example in which the vehicle-mounted camera 2 and the image processing apparatus 3 in this embodiment are mounted in a vehicle. It is a figure explaining the structure of the vehicle-mounted camera 2 and the image processing apparatus 3 in this embodiment. It is a figure explaining the edge state of a captured image. It is a figure explaining the 1st Example in this embodiment. It is a figure explaining the operation | movement procedure of the image processing apparatus 3 grade | etc., In a 1st Example. It is a figure explaining the 2nd Example in this embodiment. It is a figure explaining the 3rd Example in this embodiment. It is a figure explaining the operation | movement procedure of the image processing apparatus 3 grade | etc., In a 3rd Example. It is a figure explaining the amount of edge extraction when a focus shift is large. It is a figure explaining the 4th example in this embodiment. It is a figure explaining the operation | movement procedure of the image processing apparatus 3 grade | etc., In a 4th Example. It is a figure explaining the 5th Example in this embodiment.

Explanation of symbols

2: vehicle-mounted camera, 3 image processing device, 5: vehicle-mounted radar, 7: navigation device,
8: Control center, 32: Image processing unit, 33: Storage unit, 35: Reference value table,
71: Display unit, 72: Communication unit

Claims (7)

In an image processing apparatus that processes captured image data of a host vehicle that is imaged by a fixed-focus vehicle-mounted camera in which an imaging lens is set at a predetermined position
A storage unit that stores first characteristic data of a part of a captured image of the host vehicle captured by the in-vehicle camera when the imaging lens is at the set position;
An image processing unit that extracts second feature data of a part of the captured image of the host vehicle that is captured by the in-vehicle camera after the first feature data is stored;
The image processing unit compares the first and second feature data and determines a defocus of the in-vehicle camera based on the comparison result. In an image processing apparatus that processes captured image data around a host vehicle that is imaged by a fixed-focus-type in-vehicle camera in which an imaging lens is set at a predetermined position.
A storage unit that stores first feature data of a captured image of a subject other than the subject vehicle captured by the in-vehicle camera when the imaging lens is at the set position for each of a plurality of distances between the subject vehicle and the subject. When,
An image processing unit for extracting second feature data of a captured image of the subject imaged by the in-vehicle camera after the first feature data is stored;
The image processing unit compares the first feature data corresponding to the distance data between the subject vehicle detected by the subject distance detection unit and the subject and the second feature data, and determines the comparison result. An image processing apparatus for determining a defocus of the in-vehicle camera based on the image processing apparatus. In an image processing apparatus that processes captured image data around a host vehicle imaged by a fixed-focus vehicle-mounted camera in which an imaging lens is set at a predetermined position,
A storage unit that stores first feature data of a captured image of a subject other than the subject vehicle captured by the in-vehicle camera from a predetermined imaging point when the imaging lens is at the set position;
A second captured image of the subject captured by the in-vehicle camera when the host vehicle position detection unit detects that the host vehicle is located at the imaging point after the first feature data is stored. An image processing unit for extracting feature data of
The image processing unit compares the first and second feature data and determines a defocus of the in-vehicle camera based on the comparison result. In an image processing method for processing captured image data around a host vehicle imaged by a fixed-focus vehicle-mounted camera in which an imaging lens is set at a predetermined position,
Storing first feature data of a captured image of a part of the host vehicle captured by the in-vehicle camera when the imaging lens is at the set position;
Extracting second feature data of a captured image of a part of the host vehicle captured by the in-vehicle camera after the first feature data is stored;
And a step of comparing the first and second feature data and determining a defocus of the in-vehicle camera based on the comparison result. In an image processing method for processing captured image data around a host vehicle imaged by a fixed-focus vehicle-mounted camera in which an imaging lens is set at a predetermined position,
Storing first feature data of a captured image of a subject other than the subject vehicle captured by the in-vehicle camera when the imaging lens is at the set position for each of a plurality of distances between the subject vehicle and the subject;
Extracting second feature data of a captured image of the subject imaged by the in-vehicle camera after the first feature data is stored;
Detecting the distance between the vehicle and the subject;
Comparing the first feature data corresponding to the distance data between the host vehicle and the subject and the second feature data, and determining a defocus of the in-vehicle camera based on the comparison result. An image processing method comprising: In an image processing method for processing captured image data of a host vehicle imaged by a fixed focus type in-vehicle camera in which an imaging lens is set at a predetermined position,
Storing first feature data of a captured image of a subject other than the subject vehicle captured by the in-vehicle camera from a predetermined imaging point when the imaging lens is at the set position;
Detecting that the host vehicle is located at the imaging point after the first feature data is stored;
Extracting second feature data of a captured image of the subject from captured image data captured by the in-vehicle camera when the host vehicle is located at the imaging point;
And a step of comparing the first and second feature data and determining a defocus of the in-vehicle camera based on the comparison result. In an image processing method for processing captured image data of a host vehicle imaged by a fixed focus type in-vehicle camera in which an imaging lens is set at a predetermined position,
Transmitting the captured image data to a remote control center;
Transmitting determination result data indicating the sharpness of the captured image determined by the operator at the control center;
And a step of determining a defocus of the in-vehicle camera based on the determination result data.

Images