WO2017122688A1

WO2017122688A1 - Device for detecting abnormality of lens of onboard camera

Info

Publication number: WO2017122688A1
Application number: PCT/JP2017/000659
Authority: WO
Inventors: 大輔杉浦; 宗昭松本
Original assignee: 株式会社デンソー
Priority date: 2016-01-12
Filing date: 2017-01-11
Publication date: 2017-07-20
Also published as: JP6565693B2; JP2017126835A

Abstract

A lens abnormality detection device provided with a plurality of onboard cameras (4f, 4r), a visual point conversion unit (22), a storage unit (16), and an abnormality determination unit (26). The storage unit stores a front bird's-eye image, among bird's-eye images generated by the visual point conversion unit, that includes the front in the traveling direction of a vehicle, as a history image, successively in accordance with the traveling of the vehicle. The abnormality determination unit compares a rear bird's-eye image that includes the rear in the traveling direction of the vehicle with the history image stored in the storage unit, and determines that there is lens abnormality in one of the plurality of onboard cameras when a difference between the rear bird's-eye image and the history image is greater than or equal to an abnormality determination value.

Description

In-vehicle camera lens abnormality detection device

Cross-reference of related applications

This international application claims priority based on Japanese Patent Application No. 2016-003752 filed with the Japan Patent Office on January 12, 2016, and is based on Japanese Patent Application No. 2016-003752. The entire contents are incorporated herein by reference.

The present disclosure relates to a lens abnormality detection device that detects an abnormality of a lens of a vehicle-mounted camera.

A system is known in which an in-vehicle camera is used to capture an image of the surroundings of the vehicle so as to monitor the running state of the vehicle and notify an abnormality, or to generate a bird's-eye view of the surroundings of the vehicle as viewed from above.

In this type of system, if an abnormality such as scratches or dirt occurs on the lens of the in-vehicle camera, it becomes impossible to image the surroundings of the vehicle well, so it is possible to detect the lens abnormality by processing the captured image. Proposed.

For example, in Patent Document 1 below, two on-vehicle cameras are used to image roads ahead and behind the vehicle in the running direction, and road marking lines such as a road center line and a lane boundary line are extracted from the captured images of the on-vehicle cameras. An apparatus for calculating the reliability of road marking lines has been proposed. In this proposed device, the reliability histories of road lane markings calculated for each camera are compared, and if there is an in-vehicle camera with relatively low reliability, the lens of the in-vehicle camera is abnormal. Judge.

JP 2014-115814 A

According to the proposed apparatus, it is possible to detect a lens abnormality with higher accuracy than an apparatus that performs abnormality determination by extracting and comparing overlapping image regions from two captured images obtained by two cameras. it can. This is because if an overlapping image region is extracted from two captured images and an abnormality is determined, an abnormality of a lens portion that does not correspond to the overlapping image region (for example, dirt on the lens) cannot be detected.

However, in the proposed apparatus, road lane markings are extracted from images captured by each on-vehicle camera, and the straightness of the extracted road lane markings and the parallelism of the edges on both sides in the width direction of the road lane markings are calculated. By doing so, the reliability of the road marking line is obtained. As a result of detailed studies by the inventor, it has been found that the above-mentioned proposed apparatus cannot detect lens abnormality of an in-vehicle camera when the vehicle is traveling on a road having no road marking line.

In addition, in order to extract road lane markings from captured images and calculate their linearity and parallelism, it is necessary to perform more complicated image processing than when two images are compared. As a result of detailed examinations by the persons, it has been found that the proposed apparatus takes time to calculate the reliability of the road marking line, and thus to determine the lens abnormality.

In one aspect of the present disclosure, it is desirable that a lens abnormality of an in-vehicle camera can be accurately detected using captured images obtained from a plurality of in-vehicle cameras even when the vehicle is traveling on a road without a road marking line. .

The lens abnormality detection device for an in-vehicle camera according to one aspect of the present disclosure includes a plurality of in-vehicle cameras, a viewpoint conversion unit, a storage unit, and an abnormality determination unit.
The plurality of in-vehicle cameras capture images of the surroundings of the vehicle, and the viewpoint conversion unit generates a bird's-eye view image by converting the viewpoints of images captured by the plurality of in-vehicle cameras.

And a memory | storage part memorize | stores sequentially the front bird's-eye view image which covers the driving | running | working direction front of a vehicle among the bird's-eye images produced | generated in the viewpoint conversion part as a history image according to driving | running | working of a vehicle.
In addition, the abnormality determination unit compares the rear bird's-eye image covering the rear in the traveling direction of the vehicle among the bird's-eye images generated by the viewpoint conversion unit and the history image stored in the storage unit.

The history image used for this comparison is a history image corresponding to the rear bird's-eye view image, and the abnormality determination unit determines whether the difference between the rear bird's-eye view image and the history image is equal to or greater than the abnormality determination value. It is determined that a lens abnormality has occurred.

The lens abnormality is an abnormality of the lens that makes it impossible to capture a clear image with the in-vehicle camera, and includes, for example, scratches, damage, dirt, and the like of the lens.
Such a lens abnormality detection device determines a lens abnormality of the in-vehicle camera by comparing bird's-eye images obtained from captured images captured by a plurality of in-vehicle cameras. Therefore, in such a lens abnormality detection device, it is possible to determine lens abnormality even if the cameras that are subject to lens abnormality determination are not arranged so that the imaging regions overlap.

In addition, the lens abnormality may be determined by comparing the rear bird's-eye image and the history image corresponding to the rear bird's-eye image to obtain a difference between the rear bird's-eye image and the history image. There is no need to extract lines and calculate their linearity and parallelism. For this reason, in such a lens abnormality detection device, abnormality determination can be performed even if there is no specific target such as a road marking line on the traveling path of the vehicle.

In addition, the lens abnormality determination can be performed by comparing the bird's-eye images obtained from the respective in-vehicle cameras and obtaining the difference between the bird's-eye images. Therefore, the lens abnormality determination can be performed easily and in a short time compared to the related technology described above. It can be implemented.

It is a block diagram showing the structure of the whole imaging system of embodiment. It is explanatory drawing showing the attachment position and imaging direction of the vehicle-mounted camera to a vehicle. It is explanatory drawing showing the bird's-eye view image area | region after imaging range and viewpoint conversion with a vehicle-mounted camera. It is a flowchart showing the lens abnormality determination process performed in the image processing part of FIG. It is explanatory drawing explaining the abnormality determination operation | movement using the bird's-eye view image obtained by lens abnormality determination processing, and this.

Hereinafter, exemplary embodiments of the present disclosure will be described with reference to the drawings.
As shown in FIG. 1, the imaging system 2 of the present embodiment includes a front camera 4 f and a rear camera 4 r as vehicle-mounted cameras, a control unit 10, and a display unit 30.

As shown in FIG. 2, the front camera 4f and the rear camera 4r are attached to the front part and the rear part of the vehicle 40 so as to image the road ahead and behind the vehicle 40, respectively.
Further, in the front camera 4f and the rear camera 4r,

lenses

5f and 5r are provided on the front end side in the imaging direction, respectively.

The control unit 10 generates a bird's-eye view image of the road around the vehicle from the vertical direction, based on the images captured by the

cameras

4f and 4r. Then, the generated bird's-eye view image is displayed on the display unit 30 including a liquid crystal display or the like disposed in the vehicle interior. The control unit 10 also functions as a lens abnormality detection device of the present disclosure.

The control unit 10 includes an imaging signal input unit 12, a detection signal input unit 14, a memory 16, a display control unit 18, an image processing unit 20, and an abnormality notification unit 32.
The imaging signal input unit 12 takes in the imaging signals from the front camera 4f and the rear camera 4r and inputs them to the image processing unit 20 as captured image data.

The detection signal input unit 14 takes in detection signals from the wheel speed sensor 6 that detects the rotational speed of each wheel of the vehicle 40 and the steering angle sensor 8 that detects the steering angle of the steering, respectively, and the wheel speed data and the steering angle data. And input to the image processing unit 20.

The image processing unit 20 includes a microcomputer including a CPU, a ROM, a RAM, and the like. When the CPU executes a program in the ROM, the viewpoint conversion unit 22, the image composition unit 24, the abnormality determination unit 26, and , Function as a shadow determination unit 28.

As shown in FIG. 3, the viewpoint conversion unit 22 converts the captured image data of the front and rear of the vehicle input from the imaging signal input unit 12 from the viewpoint, thereby moving each road region in front and rear of the vehicle 40 upward. A bird's-eye image 50f and a bird's-eye image 50r viewed from the above are generated.

The viewpoint conversion unit 22 stores the bird's-eye image 50f or the bird's-eye image 50r in the memory 16 in association with the position and direction of the vehicle 40 obtained from the wheel speed data and the steering angle data representing the traveling state of the vehicle 40.

That is, when the vehicle 40 travels, the viewpoint conversion unit 22 provides a bird's-eye image covering the front in the traveling direction (hereinafter also referred to as “front bird's-eye image”) 50f or a bird's-eye image covering the rear in the traveling direction (hereinafter referred to as “rear bird's-eye image”). By storing 50r in the memory 16, a bird's-eye view image directly under the vehicle 40 that cannot be captured in real time by the

cameras

4f and 4r is stored as the history image 52.

The memory 16 is a storage unit that stores a history image, and includes a readable and writable nonvolatile memory such as an EEPROM or a flash memory.
The image composition unit 24 composes the latest bird's-

eye view images

50f and 50r generated by the viewpoint conversion unit 22 and the history image 52 stored in the memory 16 so that the vehicle 40 and the entire periphery thereof can be viewed from above. A bird's-eye view image is generated.

Then, the generated bird's-eye view image is output from the image processing unit 20 to the display control unit 18 and displayed on the display unit 30 via the display control unit 18.
The abnormality determination unit 26 compares the rear bird's-eye image 50r with the history image of the position of the vehicle 40 corresponding to the rear bird's-eye image 50r, and compares the difference between the rear bird's-eye image 50r and this history image (in other words, the degree of coincidence). ) Is calculated.

When the calculated difference is equal to or greater than the abnormality determination value (in other words, when the degree of coincidence is less than the threshold value), it is determined that an abnormality has occurred in at least one of the

lenses

5f and 5r of the front camera 4f and the rear camera 4r. Then, the occurrence of an abnormality is notified via the abnormality notification unit 32.

It should be noted that the abnormality notifying unit 32 only needs to be able to notify the passenger of the vehicle 40 of the abnormality of the

lenses

5f and 5r. For example, the abnormality notifying unit 32 may include a circuit that blinks or lights a display lamp such as an LED. A circuit for displaying an error message may be included. In addition, a voice synthesis circuit may be included so that the abnormality of the

lenses

5f and 5r can be notified by voice.

The shadow determination unit 28 determines whether or not the shadow of the vehicle 40 is reflected in at least one of the rear bird's-eye view image and the history image used by the abnormality determination unit 26 for the abnormality determination of the

lenses

5f and 5r. When the shadow determination unit 28 determines that one of the rear bird's-eye view image and the history image has a shadow of the vehicle 40, the abnormality determination unit 26 prohibits the abnormality determination of the

lenses

5f and 5r.

This is because, when the shadow of the vehicle 40 is reflected in one of the rear bird's-eye view image and the history image used for determining the abnormality of the

lenses

5f and 5r, when these images are compared, the difference between these images increases due to the influence of the shadow. This is because it is considered that the abnormality of the

lenses

5f and 5r is erroneously determined.

Next, lens abnormality determination processing executed in the image processing unit 20 to realize functions as the viewpoint conversion unit 22, the abnormality determination unit 26, and the shadow determination unit 28 will be described with reference to the flowchart of FIG. To do.

The lens abnormality determination process shown in FIG. 4 represents a program that is repeatedly executed as one of the main routines in the microcomputer (specifically, the CPU) included in the image processing unit 20.

In the present embodiment, this program is stored in the ROM in the image processing unit 20. However, for example, other non-transitional tangible records separate from the microcomputer included in the image processing unit 20 such as the memory 16. It may be stored on a medium.

As shown in FIG. 4, in the lens abnormality determination process, first, in S110 (S represents a step), images captured by the front camera 4f and the rear camera 4r are captured from the imaging signal input unit 12, and the front bird's-eye image 50f and the rear are captured. A bird's-eye view image 50r is generated.

The process of S110 is a process for realizing the function as the viewpoint conversion unit 22. In S110, the viewpoint image of the captured image is converted to generate a bird's-eye view image. This generation method is publicly known and, for example, a technique described in Japanese Patent Laid-Open No. 10-211849 can be used. Detailed description is omitted.

Next, in S120, it is determined whether or not the vehicle 40 is currently traveling by fetching the wheel speed data from the detection signal input unit 14. Then, if the vehicle 40 is not traveling, the lens abnormality determination process is temporarily terminated. If the vehicle 40 is traveling, the process proceeds to S130.

In S130, the traveling direction (that is, forward or backward) of the vehicle 40 is determined from the wheel speed data, and the front bird's-eye image 50f is selected from the bird's-

eye images

50f and 50r generated in S110. Then, the selected front bird's-eye image 50f is stored in the memory 16 as a history image during vehicle travel.

The latest bird's-

eye images

50f and 50r generated in S110 and the history image stored in the memory 16 in S130 are displayed as bird's-eye images around the host vehicle by display control processing as the image composition unit 24. It is used for displaying on the unit 30.

Next, in S140, the rear bird's-eye image 50r is selected from the bird's-

eye images

50f and 50r generated in S110, and it is determined whether or not there is a shadow of the vehicle 40 in the selected rear bird's-eye image 50r.

If it is determined in S140 that the shadow of the host vehicle 40 is present in the rear bird's-eye view image 50r, it may be erroneously determined that the

lens

5f or 5r is abnormal in the process described later. The process is temporarily terminated.

On the other hand, if it is determined in S140 that there is no shadow of the host vehicle 40 in the rear bird's-eye image 50r, the process proceeds to S150, and the history image corresponding to the rear bird's-eye image 50r selected in S140 is read from the memory 16. .

That is, the front bird's-eye view image 50f is sequentially stored in the memory 16 as a history image while the vehicle 40 is traveling. Therefore, in S150, based on the vehicle speed obtained from the wheel speed data, the front bird's-eye image 50f at the same position as the rear bird's-eye image 50r is extracted from the history image stored in the memory 16.

In subsequent S160, it is determined whether or not the history image read from the memory 16 in S150 has a shadow of the host vehicle 40.
If it is determined in S160 that there is a shadow of the host vehicle 40 in the history image, it is conceivable that the abnormality of the

lenses

5f and 5r is erroneously determined in the process described later. finish. If it is determined in S160 that there is no shadow of the host vehicle 40 in the history image, the process proceeds to S170.

Here, the processing of S140 and S160 is processing for realizing the function as the shadow determination unit 28. In these processes, a determination method for determining whether or not there is a shadow in the bird's-eye view image is known, and for example, a shadow detection technique described in Japanese Patent Application Laid-Open No. 2010-237976 can be used.

That is, in these processes, for example, a region of a partial region of the bird's-eye view image is divided based on the hue and the luminance, and the difference in hue is equal to or less than a predetermined threshold among the plurality of divided regions. Two areas where the difference in brightness is equal to or greater than a predetermined value are extracted.

Of the two extracted areas, the high-brightness area is the non-shadow area, the low-brightness area is the shadow area, and the vector from the shadow area to the non-shadow area in the color information space is specified as the color information of the light source. To do.

Next, for all the regions of the bird's-eye view image, region division is performed based on the hue and brightness, and when the hue difference between the adjacent regions matches the hue of the light source within a predetermined range, The low brightness area is identified as a shadow.

In S140 and S160, it is determined whether the shadow area thus identified corresponds to the vehicle body shape of the host vehicle 40, so that the rear bird's-eye view image 50r or the history image includes the shadow of the host vehicle 40. Can be determined.

Japanese Patent Laid-Open No. 2010-237976 introduces a technique for detecting shadows by a procedure different from the above procedure. In S140 and S160, the shadow detection technique is used to detect the shadow of the host vehicle 40. It may be determined, or another shadow detection technique may be used.

In addition, even if the rear bird's-eye view image 50r includes a shadow of the host vehicle 40, if the shadow area is small and does not affect the abnormality determination of the

lenses

5f and 5r, the host vehicle 40 determines in S140 and S160. You may make it judge that the shadow of the vehicle 40 does not exist.

Next, in S170, the difference between the rear bird's-eye view image 50r determined to have no shadow of the host vehicle 40 in S140 and S160 and the history image is calculated. This difference calculation is performed by, for example, extracting feature points from the rear bird's-eye view image 50r and the history image by edge detection or the like, and comparing the feature points with each other, so that the difference between the rear bird's-eye view image 50r and the history image (in other words, This is done by calculating the degree of coincidence.

In the subsequent S180, the difference calculated in S170 is compared with a preset abnormality determination value. If the difference is less than the abnormality determination value, the images captured by the front camera 4f and the rear camera 4r are both normal. If it is determined that there is, the lens abnormality determination process is temporarily terminated.

In S180, when the difference is equal to or larger than the abnormality determination value, one of the images captured by the front camera 4f and the rear camera 4r is abnormal, and one of the

lenses

5f and 5r of each camera is abnormal such as breakage or dirt. The process proceeds to S190.

Note that the processing of S170 and S180 is processing for realizing the function as the abnormality determination unit 26.
In S190, by prohibiting execution of the display control process as the image composition unit 24, the generation of the bird's-eye image around the host vehicle 40 and the output of the bird's-eye image to the display control unit 18 are prohibited. The display of the bird's-eye view image on 30 is stopped.

Finally, in S200, the abnormality notification unit 32 notifies the occupant of the abnormality of the

lenses

5f and 5r, and the lens abnormality determination process is temporarily terminated. Since the lens abnormality determination process is executed as one of main routines in the CPU, the process of S110 is started after a predetermined time has elapsed.

As described above, in the imaging system 2 of the present embodiment, as illustrated in FIG. 5, when the vehicle 40 is traveling in the forward direction, the front bird's-eye image 50 f generated from the image captured by the front camera 4 f. Are sequentially stored in the memory 16 as a history image.

The history image stored in the memory 16 is used to generate a bird's-eye image around the vehicle 40 together with the latest bird's-

eye images

50f and 50r generated from images captured by the front camera 4f and the rear camera 4r.

Further, as the vehicle 40 travels, a bird's-eye image in the same area as the bird's-eye image 50f (t1) stored in the memory 16 as a history image at time T1 is generated from an image captured by the rear camera 4r at time T2. .

In this embodiment, the bird's-eye view image 50r (t2) generated at time T2 is compared with the bird's-eye view image 50f (t1) that is a history image corresponding to the bird's-eye view image 50r (t2), and the difference between the two images is compared. Is greater than or equal to the abnormality determination value, it is determined that the

lens

5f or 5r is abnormal.

Therefore, as illustrated in FIG. 5, if the bird's-eye image 50r (t2) and the bird's-eye image 50f (t1) as the history image include the sign 60 representing a pedestrian crossing, the edge of the sign 60 is detected from each image. Then, the abnormality of the

lenses

5f and 5r is determined from the difference in position and shape.

Even if the sign 60 is not shown in each image, the road surface pattern and unevenness, the color difference at the boundary between the traveling road and its surroundings, and the like are extracted as feature points by edge detection, and the lens is determined from these differences. It is determined whether or not there is an abnormality in 5f and 5r.

Therefore, according to the imaging system 2 of the present embodiment, it is not necessary to arrange the cameras that are the targets of abnormality determination so that the imaging areas overlap. Further, as described in Patent Document 1, it is not necessary to extract road lane markings from images captured by the front camera 4f and the rear camera 4r and calculate their linearity and parallelism.

For this reason, according to the present embodiment, by mounting a plurality of

cameras

4f and 4r on the vehicle 40, these

cameras

4f and 4f, even if there is no specific target such as a road marking line on the road surface on which the vehicle 40 travels, The abnormality of the

4r lenses

5f and 5r can be respectively determined.

In addition, this abnormality determination only needs to obtain a difference between the bird's-eye images by comparing the bird's-eye images at the same position of the vehicle 40 generated from the captured images of the

cameras

4f and 4r. Compared to the case of calculating linearity and parallelism, it can be carried out easily and in a short time.

In the present embodiment, when one of the rear bird's-eye view image and the history image used for abnormality determination of the

lenses

5f and 5r includes a shadow of the own vehicle, the abnormality determination is not performed.

Therefore, according to the present embodiment, for example, when a shadow is formed on the front side or the rear side of the vehicle 40 due to sunlight, the abnormality of the

lenses

5f and 5r is erroneously determined due to the influence of the shadow. Can be prevented or suppressed.

Further, in the present embodiment, when the abnormality of the

lenses

5f and 5r is detected, the display of the bird's-eye image on the display unit 30 by the image composition unit 24 is prohibited, and the abnormality of the

lenses

5f and 5r is detected via the abnormality notification unit 32. Inform. For this reason, when an abnormality such as breakage or dirt occurs in the lens of the front camera 4f or the rear camera 4r, and a bird's-eye image around the vehicle 40 cannot be normally generated, a blurred bird's-eye image is displayed on the display unit 30. Can be suppressed. In addition, by notifying the abnormality of the

lenses

5f and 5r, it is possible to prompt the user to check each of the

cameras

4f and 4r.

As mentioned above, although one embodiment of this indication was described, this indication is not limited to the above-mentioned embodiment, and can carry out various modifications.
For example, although the above embodiment has been described on the assumption that the front camera 4f and the rear camera 4r are mounted on the vehicle 40, the present disclosure is not limited to the above embodiment if a plurality of cameras are mounted on the vehicle 40. Lens abnormality can be determined in the same manner as above.

In other words, in addition to the front camera 4f and the rear camera 4r, the vehicle 40 may be equipped with side cameras that capture the left and right sides of the vehicle, respectively.
In this case, a front bird's-eye image and a rear bird's-eye image of the vehicle are generated using the front camera and one side camera, and the front bird's-eye image is left as a history image. Then, the lens abnormality is determined by comparing the image regions having the same position in the rear bird's-eye view image and the history image.

Even in this case, the lens abnormality of each camera can be determined in the same manner as in the above embodiment. Similarly, a lens abnormality can be determined using the rear camera and one side camera.

In this case, one or both of the front camera 4f, the rear camera 4r, and the side camera are used to generate bird's-eye images, and image regions whose positions match are extracted from the generated bird's-eye images and compared. By doing so, lens abnormality may be determined.

Next, in the above embodiment, the image processing unit 20 includes a microcomputer, and the CPU of the microcomputer executes the program, whereby the viewpoint conversion unit 22, the image composition unit 24, the abnormality determination unit 26, and the shadow determination unit 28. As a function is realized.

However, the image processing unit 20 is configured to implement part or all of the functions using hardware that combines a logic circuit, an analog circuit, or the like in place of or in addition to the microcomputer. May be.

In the above-described embodiment, the wheel speed sensor 6 and the steering angle sensor 8 are described as being used so that the position and traveling direction of the vehicle 40 can be grasped. However, a satellite positioning system is used separately from these. You may do it.

Next, a plurality of functions of one constituent element in the above embodiment may be realized by a plurality of constituent elements, or one function of one constituent element may be realized by a plurality of constituent elements. . Further, a plurality of functions possessed by a plurality of constituent elements may be realized by one constituent element, or one function realized by a plurality of constituent elements may be realized by one constituent element. Moreover, you may abbreviate | omit a part of structure of the said embodiment. Further, at least a part of the configuration of the above embodiment may be added to or replaced with the configuration of the other embodiment. In addition, all the aspects included in the technical idea specified only by the wording described in the claims are embodiments of the present disclosure.

In addition to the imaging system 2 described above, the present disclosure can be realized as a system including the imaging system 2 as a component, or a program for causing a computer to function as the image processing unit 20 of the imaging system 2. it can. Further, the present disclosure can be realized in various forms such as a non-transitional actual recording medium such as a semiconductor memory in which the program is recorded, a lens abnormality detection method realized by the image processing unit 20, and the like.

Claims

A plurality of in-vehicle cameras (4f, 4r) configured to image the surroundings of the vehicle;
A viewpoint conversion unit (22, S110) configured to generate a bird's-eye view image by converting viewpoints of images captured by the plurality of in-vehicle cameras,
A storage unit configured to sequentially store, as a history image, a front bird's-eye image covering the front in the traveling direction of the vehicle among the bird's-eye images generated by the viewpoint conversion unit. (16, S130),
Among the bird's-eye images generated by the viewpoint conversion unit, a rear bird's-eye image covering the rear in the traveling direction of the vehicle is compared with a history image stored in the storage unit and corresponding to the rear bird's-eye image. When the difference between the rear bird's-eye view image and the history image is greater than or equal to an abnormality determination value, an abnormality determination unit configured to determine that a lens abnormality has occurred in one of the plurality of in-vehicle cameras (26, S160-S180),
Lens abnormality detection device for in-vehicle camera equipped with
A shadow determination unit (28, S140, S150) configured to determine whether or not the bird's-eye image generated by the viewpoint conversion unit includes a shadow of the host vehicle;
The abnormality determination unit is configured to prohibit the lens abnormality from being determined using a bird's-eye image that is determined by the shadow determination unit to include the shadow of the host vehicle. Item 2. An on-vehicle camera lens abnormality detection device according to Item 1.
An image synthesis unit (24) configured to generate a bird's-eye image around the host vehicle by synthesizing the bird's-eye image generated by the viewpoint conversion unit and the history image stored in the storage unit. )
The in-vehicle camera lens abnormality detection device according to claim 1, wherein the abnormality determination unit is configured to prohibit image composition by the image composition unit when it is determined that the lens abnormality has occurred.
An abnormality notifying unit (32, S190) configured to notify that the lens abnormality has occurred when the abnormality determining unit determines that the lens abnormality has occurred. The lens abnormality detection device for an in-vehicle camera according to any one of claims 1 to 3.