JP2008060874A - On-vehicle camera and stain detector for on-vehicle camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for detecting waterdrops and stains adhered to optical components of an imaging apparatus, in the imaging apparatus that has been installed outside the room of a vehicle. <P>SOLUTION: Optical flow of prescribed pixels is measured from a plurality of images picked up at different timing with the imaging apparatus, and the optical flow is further deduced from the vehicle speed information and information from a car navigation system. The presence or the absence of the waterdrops or stains, adhered to the optical components of the imaging apparatus, is decided based on comparison results between the optical flow deduced and the measured optical flow. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an in-vehicle camera, and more particularly to an apparatus and a detection method for detecting water droplets and dirt attached to an imaging apparatus attached outside a vehicle such as an automobile.

  As a technology to improve convenience and safety in automobiles, a camera is installed in the vehicle, and white lines and road markings on the road are recognized from images captured by the camera, obstacles behind the vehicle are detected, and parking is performed. Devices for assisting are known.

  A camera used in such an apparatus is assumed to be attached to the outside of the vehicle, and there is a possibility that water droplets or dirt may adhere to the lens of the camera. Depending on the degree of water droplets and dirt adhering to the lens, it may be impossible to satisfy the convenience function and the safety function described above. Therefore, there is a need for means for detecting water droplets and dirt adhering to the lens of the camera, giving a warning to the user, and automatically cleaning.

  In Patent Document 1, when a vehicle is moving, the difference between density values included in images taken by two cameras at different timings is accumulated over a plurality of frames, and based on the accumulated density value images. The camera is detecting dirt on the camera.

  In Patent Document 2, a difference value between the amount of solar radiation detected by the vehicle air conditioner solar radiation sensor installed in the vehicle interior and the amount of solar radiation calculated based on the video information from the camera is calculated and further calculated. When the difference value is integrated, the difference value is calculated and integrated (N-1) times, and the difference value integrated for the nth time exceeds a predetermined difference value, the optical system component of the camera is contaminated. Deciding.

JP 2003-259358 A JP 2005-198004 A

  In the method described in Patent Document 1, the difference between density values included in images taken by two cameras at different timings is accumulated over a plurality of frames, and the image is attached to the camera based on the accumulated density value images. Detected dirt. For this reason, it is difficult to detect a moving deposit such as a water droplet.

  In addition, it is effective when the background is hidden by dirt adhering to the lens (such as mud dirt), but it is difficult to detect adhering substances that easily pass through the background, such as water droplets, unless the accumulated time is increased. , It takes time to detect dirt.

  In the method described in Patent Document 2, dirt is removed based on a difference value between the amount of solar radiation detected by a vehicle air conditioner solar radiation sensor installed in a vehicle interior and the amount of solar radiation calculated based on video information from the camera. Detected.

  Since the location of the solar radiation sensor and the camera is usually different, an error may occur between the solar radiation amount at the solar radiation sensor and the solar radiation amount calculated from the camera (for example, when the solar radiation sensor part has a shadow).

  Also, at night, the solar radiation sensor hardly detects the amount of solar radiation, but there is a possibility that light from oncoming vehicles, preceding vehicles, street lamps, etc. enters the image captured by the camera and is calculated as the amount of solar radiation. Therefore, an error occurs in the amount of solar radiation calculated from both sensors, and there is a possibility that dirt cannot be detected accurately.

  In order to solve the above-described problems, the present invention provides an imaging unit that includes an optical system and an imaging device and images a vehicle exterior, and a movement amount of a predetermined extract from a plurality of images captured at different timings by the imaging device. And an optical flow measuring means for calculating the moving direction, and an optical flow estimating means for estimating the moving amount and the moving direction of the extract based on the vehicle speed information, and the optical flow measuring means measured by the optical flow measuring means. Based on the movement amount and movement direction of the extract and the movement amount and movement direction of the extract estimated by the optical flow estimation means, the adhering matter attached to the optical system of the imaging means is detected.

  The adhering matter adhering to the optical system of the in-vehicle camera can be detected in a shorter time than in the prior art. Further, it is possible to detect a deposit that moves relative to the optical system, such as a water droplet, or a deposit that easily passes through the background. Furthermore, practical adhesion detection can be performed without using another sensor such as a sunshine sensor.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 shows an example of an on-vehicle camera deposit detection apparatus 9 according to the present invention.

  In this embodiment, the on-vehicle camera deposit detection device 9 includes an A / D conversion circuit 2 that converts an analog image signal acquired by the imaging device 1 into a digital image signal, and an image processing device that processes the digital image signal. 3, a memory 4, and an arithmetic device 5.

  The computing device 5 is connected to the car navigation system 6 and can acquire map information such as a road shape. The computing device is connected to a vehicle network, and information can be acquired from other sensors and control devices mounted on the vehicle 10 via the vehicle network. For example, a vehicle speed sensor is connected to the vehicle network, and vehicle speed information can be acquired. Hereinafter, information obtained from such other sensors and control devices is referred to as vehicle information 7. Further, the arithmetic device 5 is also connected to the alarm device 8, and can alert the user when there is a deposit 80 on the optical system component 13 of the imaging device 1. The on-vehicle camera adhering matter detection device 9 and the imaging device 1 may be configured as an on-vehicle camera housed in the same casing, and may be separately controlled to be connected to the imaging device 1 via a communication line. You may comprise as a function of an apparatus.

  As shown in FIG. 2, in the present embodiment, the imaging device 1 is installed in the rear part of the vehicle 10 and images the rear of the vehicle. The angle of view of the imaging device 1 is 45 degrees. The imaging range 11 is set to infinity from the bumper of the own vehicle.

  Next, the operation of the in-vehicle camera deposit detection apparatus 9 according to the present invention will be described with reference to the flowchart of FIG.

  First, initial setting is performed in step S1. In the initial setting, each parameter necessary for recognition is initialized.

  Next, vehicle information 7 is acquired at Step S2. The vehicle information 7 includes vehicle speed information, yaw rate sensor information, and steering angle information from the vehicle network.

  In step S3, information from the car navigation system 6 is acquired. From the car navigation system 6, map information around the host vehicle is acquired. This map information includes the curvature of the road on which the vehicle is traveling, the width of the road, and the type information of the white line 61 and road marking 60 drawn on the road. Moreover, the request | requirement of the deposit | attachment detection process from the car navigation system 6 is included.

  Subsequently, in step S4, based on the vehicle information 7 acquired in steps S2 and S3 and the information from the car navigation system 6, it is determined whether or not the attached matter detection process is executed.

  When the vehicle speed acquired in step S2 is equal to or higher than a predetermined speed, and when the white line information obtained in step S3 and the type information of the road marking 60 are obtained, or the adhering matter detection processing from the car navigation system 6 If there is a request, the process proceeds to the next step S5 to execute the adhering matter detection process. In the case of other conditions, the attached matter detection process is not executed, and the process returns to step S2. The predetermined speed may be a lower limit value at which the optical flow can be calculated based on a plurality of images captured at different timings. Here, the optical flow is a vector indicating in what direction and how far a certain point or figure in the image moves at the next moment, and is captured at different timings as described later. It can be obtained from the difference between the plurality of images.

  Therefore, as the above-mentioned predetermined speed, the adhering matter detection process should not be executed theoretically only when the vehicle speed is 0 km / h. However, in practice, it is necessary to consider the error of the vehicle speed sensor. A wheel speed sensor frequently used as a vehicle speed sensor has a characteristic that an error becomes large in a low speed region. Therefore, it is desirable that the adhering matter detection is not performed below a predetermined speed (for example, 5 km / h or 10 km / h).

  In step S5, the processing target area in the acquired image is set based on the vehicle information 7 acquired in steps S2 and S3 and the information from the car navigation system 6. In the case of the present embodiment, the magnitude of the optical flow varies depending on the magnitude of the vehicle motion. For example, when images are taken at the same time interval, the position where the same object appears on a plurality of images is greatly shifted when the vehicle speed is high. Accordingly, when the vehicle motion is small, the size of the optical flow is predicted to be relatively small. Therefore, it is not efficient to perform the optical flow measurement process on the entire imaging region. Therefore, in this embodiment, the processing area for optical flow measurement is set according to the vehicle speed, the steering angle, the yaw rate, or a combination thereof.

  The setting of the processing area includes the measurement range 21 of the optical flow 50, the position and number of the measurement points 20 of the optical flow 50, and the search range 40 at each set optical flow measurement point 20.

  When the vehicle speed acquired in step S2 is smaller than the predetermined speed, or when the vehicle speed and the steering angle are smaller than the predetermined magnitude, as shown in FIG. 3 (a), the ranges are 0.5m, 1m, and 2m. The measurement point 20 of the optical flow 50 is set.

  In the case of the above conditions, the search area 40 at each measurement point 20 in the optical flow 50 is set narrow as shown in FIG. In this embodiment, the vicinity of 8 with respect to the measurement point 20 is set as the search area 40.

  On the other hand, when the vehicle speed acquired in step S2 is larger than a predetermined speed, when the yaw rate is larger than a predetermined magnitude, or when the vehicle speed and the steering angle are larger than a predetermined magnitude, as shown in FIG. The measurement point 20 of the optical flow 50 is set in the range of 0.5 m, 1 m, 2 m, and 3 m.

  In the case of the above conditions, the search area 40 at each measurement point 20 of the optical flow 50 is set to be wide as shown in FIG. In this embodiment, the vicinity of 24 with respect to the measurement point 20 is set as the search area 40.

  Also, considering the processing load, the density of the measurement points 20 of the optical flow 50 is thinned out to reduce the number of measurement points. Further, the search area 40 at the measurement point 20 of each set optical flow 50 is set wide.

  As described above, the measurement range 21 and the number of measurement points 20 of the optical flow 50 and the search area 40 are variable based on at least one of the vehicle speed, the steering angle, and the yaw rate. As a result, the optical flow 50 can be measured with high accuracy, and the arithmetic device 5 can be used efficiently.

  In step S6, an estimated value 70 of the optical flow 50 is calculated based on the vehicle information 7 obtained in steps S2 and S3 and the information from the car navigation system 6.

  Optical flow measured from the white line 61 drawn on the road acquired from the car navigation system 6 and the type information of the road marking 60, the vehicle speed information acquired from the vehicle information 7, the yaw rate information, and the steering angle information. Estimate the size and angle of 50.

  Since the size and pattern of the white line 61 and the road marking 60 obtained from the car navigation system 6 are known, the optical flow 50 at each measurement point 20 can be accurately estimated from the information shown above.

  In addition, since the mounting angle of the imaging device 1 with respect to the vehicle 10 is substantially constant, each pixel on the imaging element has a one-to-one relationship between the distance and azimuth from the own vehicle of the area shown in the pixel. ing. Therefore, if information on the vehicle speed, rudder angle or yaw rate, or a combination thereof is obtained, what is reflected in each pixel of the captured image at a certain timing in the situation where the speed, rudder angle, and yaw rate are generated. It is possible to estimate where the captured image appears at the next timing. Therefore, the optical flow can be estimated by referring to the map or performing coordinate conversion from the vehicle speed, the steering angle, the yaw rate, etc. without using the image information actually captured.

  Next, in the image acquisition setting in step S7, the image capture period, the shutter speed, and the gain of image brightness are set based on the vehicle information 7 obtained in steps S2 and S3 and the information from the car navigation system 6. To do.

  If the vehicle speed acquired in step S2 is greater than a predetermined speed, or if the yaw rate is greater than a predetermined magnitude, or if the vehicle speed and the steering angle are greater than a predetermined magnitude, the image capture cycle is set short, Set the shutter speed faster.

  Thereby, even when the moving speed of the vehicle 10 is fast, an image without image blur can be acquired, and the optical flow 50 can be accurately measured.

  Further, the shutter speed and gain of the imaging device 1 are set based on the luminance distribution of the captured image. Since the surroundings are dark at night or in a tunnel, it may be difficult to capture a clear image of the white line 61 and the road marking 60 drawn on the road.

  In this case, the shutter speed is decreased or the gain is increased, and each is set so that the white line 61 and the road marking 60 drawn on the road can be reliably imaged.

  Subsequently, the image (image ID: n) captured by the imaging device 1 is stored in the memory 4 in the image capturing process in step S8.

  Next, in step S9, the optical flow 50 is measured. The optical flow 50 can be measured by comparing the image ID = n−1 and the image ID = n acquired at a predetermined cycle (n is a natural number).

  To which point in the image of image ID = n the density value of the measurement point 20 of the optical flow 50 in the image ID = n−1 is moved with respect to the search area 40 of each pixel set in step S5. Explore. In the example shown in FIG. 5A and FIG. 5B, a specific feature (extract) shown in the center pixel in the image with ID = n−1 is shown in the upper right corner in the image with ID = n. Has moved. Therefore, by comparing the two, the size of the optical flow 50 (the amount of movement of the extract on the image) and the angle (similarly, the direction of movement on the image) are calculated from the coordinates of the pixel searched for ID. The

  In addition, since the spatiotemporal gradient method, the collation method, etc. already exist as a well-known technique in the measurement method of the optical flow 50, detailed description is abbreviate | omitted in a present Example.

  Next, in the optical flow analysis process in step S10, the optical flow 70 estimated in step S6 and the optical flow 50 measured in step S9 are collated.

  The magnitude and angle of the optical flow 70 at each measurement point 20 estimated at step S6 are compared with the optical flow 50 at each measurement point 20 measured at step S9, and a correlation value is calculated for each measurement point 20.

  Here, a scene is assumed in which the vehicle passes through a diamond-shaped road marking 60 as shown in FIG.

  First, as shown in FIG. 7, the optical flow 70 estimated in step S6 has a size and an angle that depend on the behavior of the vehicle 10, the road marking 60 drawn on the road, and the road shape.

  When there is no deposit 80 on the optical system component 13 of the imaging apparatus 1, the optical flow 50 measured in step S9 is measured as a size and an angle that are substantially the same as the estimated value as shown in FIG. Therefore, a high correlation value is obtained.

  On the other hand, when there is a deposit on the optical system component 13 of the imaging apparatus 1, the optical flow 50 measured in step S9 has a measurement point different from the estimated value as shown in FIG.

  When there is no deposit 80, the optical flow 50 is measured by the road surface texture (white line 61 drawn on the road, road marking 60, etc.), but the road surface texture is measured by the deposit 80 such as water drops or dirt. Cannot be measured accurately.

  Therefore, the vehicle behavior estimated from the information obtained in step S2 and the optical flow 70 estimated from the white line 61 and road marking 60 drawn on the road obtained in step S3 are different in size and angle. The optical flow 50 is measured. Therefore, the correlation value calculated in step S10 is a low value.

  Next, in step S11, the presence or absence of the deposit 80 is determined based on the correlation value calculated in step S10. That is, when the correlation value is greater than or equal to a predetermined value, it is determined that there is no deposit (step S12), and when the correlation value is lower than the predetermined value, it is determined that there is an deposit (step S13).

  If it is determined in step S11 that there is a deposit, an alarm signal is output in step S14.

  The alarm signal in step S11 is notified to the user by outputting it to an alarm device 7 such as a warning lamp. Further, it may be output to the display screen of the car navigation system 6 or warned with a warning sound or voice.

  As described above, an application 80 (for example, white line recognition, marker recognition on the road, etc.) using the imaging device 1 by automatically detecting the adhering matter 80 such as water droplets and dirt attached to the optical system component 13 of the imaging device 1. It is possible to warn in advance of deterioration in performance of vehicle detection and obstacle detection functions.

  By automatically detecting the adhering matter 80 such as water droplets and dirt adhering to the optical components of the image pickup apparatus 1, the inspection that the user has regularly performed becomes unnecessary and efficient maintenance becomes possible.

  In this embodiment, the installation position of the imaging device 1 is the rear part of the vehicle, but it may be installed on the side part or the front part of the vehicle. In addition, although the angle of view of the lens in the imaging apparatus 1 is 45 degrees, the present invention can be applied to a lens having an arbitrary angle of view in accordance with a target application.

  In this embodiment, the vehicle speed information is acquired from the vehicle network, but GPS information in the car navigation system 6 may be used. Further, in the yaw rate information, information from a yaw rate sensor provided in the car navigation system 6 may be used.

  In this embodiment, the vehicle speed information and the information from the car navigation system 6 are used to determine whether or not the attached matter detection process is executed, but information on the transmission of the vehicle may be used. For example, when the transmission is in the parking position, the attached matter detection process may not be executed.

  In this embodiment, the vehicle speed information and the information from the car navigation system 6 are used to determine whether or not the attached matter detection process is executed, but information on the wiper operation of the vehicle may be used. For example, when the wiper is operating, there is a high possibility that the wiper is in a rainy state, and thus there is a high possibility that water droplets are attached to the optical components of the imaging device 1. Therefore, the attached matter detection process may be executed when the wiper is operating.

  Moreover, although the arithmetic device 5 of the on-vehicle camera deposit detection device 9 of this embodiment is different from the car navigation system 6, the arithmetic device 5 of the car navigation system 6 may be used in common. In this case, the presence or absence of the attached matter detection process may be determined based on the load factor of the arithmetic device 5. For example, in the car navigation system 6, the calculation load of the car navigation system 6 increases during map search and map drawing. In this case, the attached matter detection process is not executed, and the process of the car navigation system 6 can be prioritized.

  In this embodiment, the measurement point 20 of the optical flow 50 is set to 0.5 m, 1 m, 2 m, and 3 m, but can be set to any position in the imaging range 11.

  In the present embodiment, the search area 40 at each measurement point 20 of the optical flow 50 is set to 8 and 24 in accordance with the vehicle speed. However, the number of pixels of the image sensor, the focal length of the lens, and the image in the image capturing apparatus 1 It can be arbitrarily changed according to the image size used for processing.

The basic embodiments of the present invention have been described above. However, in addition to the above-described embodiments, configurations as listed below can be adopted. Moreover, the effect demonstrated below is acquired by these structures.
[1] In an apparatus for detecting adhering matter 80 such as water droplets and dirt adhering to the optical system component 13 of the imaging apparatus 1 installed outside the vehicle, a plurality of images captured at different timings by the imaging apparatus 1 are used. Optical flow measuring means for calculating a moving amount and moving direction of a predetermined pixel, and optical flow estimating means for estimating the optical flow 50 from the vehicle speed information.

  The movement amount and movement direction of the predetermined pixel measured by the optical flow measurement means are compared with the movement amount and movement direction of the predetermined pixel estimated by the optical flow estimation means. Based on this comparison result, the adhering matter 80 such as water droplets and dirt adhering to the optical system component 13 of the imaging apparatus 1 is detected.

As a result, it is possible to detect an adhering substance that easily passes through the background, such as a water drop, without taking a processing time as compared with the prior art. In addition, it is possible to detect even a moving deposit 80, such as a water drop, with higher accuracy than in the prior art. It is possible to accurately detect the deposit 80 without using a sunshine sensor.
[2] In addition to the vehicle speed information 7, yaw rate sensor information, rudder angle sensor information, GPS information, and map information of the car navigation system 6 may be used in combination for the optical flow estimation means.

By using the information other than the vehicle speed information, the behavior of the vehicle 10 can be estimated, so that the estimation accuracy of the optical flow 50 is improved, and adhered matter such as water droplets and dirt adhered to the optical system component 13 of the imaging device 1. 80 can be detected more reliably. Further, since the optical flow 70 can be estimated from the movement of the vehicle 10, it is possible to accurately determine whether the deposit 8 such as water droplets or dirt attached to the optical system component 13 of the imaging apparatus 1 or the structure such as the paint 60 or the building on the road surface. Can be distinguished well.
[3] In the in-vehicle camera adhering matter detection device 9, the search area 40 of the optical flow 50 in the optical flow measuring means is narrowed or optical flow measurement is performed according to the optical axis direction and the angle of view of the imaging device 1. The coordinates of the point 20 may be changed. This makes it possible to measure the optical flow 50 in the imaging range without unevenness.
[4] In the present embodiment, communication with the car navigation system 6 may be performed, the deposit 80 may be detected based on communication information from the car navigation system 6, and the processing sequence may be changed. Specifically, when information indicating that a road marking paint 60 (for example, a pedestrian crossing or a deceleration zone) is present in the imaging range 11 of the in-vehicle camera is received by communication information with the car navigation system 6, the optical flow is received. The measuring means is executed.

  Thereby, when there is no road marking paint 60, since the measurement process of the optical flow 50 is not performed, the arithmetic unit 5 can be operated efficiently.

Furthermore, since the shape and color of the road marking paint 60 are known, the optical flow 50 can be measured more accurately. Therefore, it is possible to accurately detect the deposit 80 such as water droplets or dirt adhering to the optical system component 13 of the imaging device 1.
[5] In the optical flow measurement means, when the vehicle speed calculated from the vehicle speed sensor or GPS information is high, the yaw rate of the vehicle 10 is large, the steering angle is large, and the like, the optical flow 50 search area 40 and the measurement are measured. The range 21 can be narrowed. With this configuration, it is possible to adjust the amount of calculation required for the measurement of the optical flow 50 according to the state of the vehicle motion, and to operate the calculation device 5 more efficiently. Also, it is desirable to reduce the number of optical flow measurement points 20 under the above conditions.
Note that the search area 40 and the measurement range 21 of the optical flow 50 may be narrowed even when the load on the arithmetic device 5 is high regardless of the state of the vehicle motion. Thereby, the processing load of the arithmetic unit 5 can be reduced and the arithmetic efficiency of the entire system can be increased.
[6] The frequency of executing the attached matter detection process may be changed according to the vehicle speed information. As a result, the calculation load can be further improved in efficiency. Further, in the means for changing the execution frequency of the attached matter detection processing, when it is predicted that the load factor of the computing device 5 is increased by a load factor of the computing device 5, that is, processing other than the attached matter detection processing, The execution frequency of the deposit detection process can be reduced. further,
[7] In this embodiment, when it is considered that the possibility of raindrops being attached is low, the frequency of executing the attached matter detection process may be reduced. Specifically, the arithmetic device 5 receives the wiper information as the vehicle information 7, and when the wiper is not operating, the frequency of executing the attached matter detection process is lowered. This enables smooth system operation.
[8] In this embodiment, the exposure time of the image sensor may be controlled based on the luminance distribution of the captured image, or the gain for the charge accumulated in the image sensor may be controlled. Thereby, the deposit | attachment 80 adhering to the optical system component 13 of the imaging device 1 can be detected accurately also at night.

It is a figure which shows an example of the adhesion detection apparatus 9 for vehicle-mounted cameras concerning this invention. It is a figure which shows an example of the installation position and imaging range 11 of the adhesion detection apparatus 9 for vehicle-mounted cameras. It is a figure which shows an example of the measurement point 20 and the measurement range 21 of the optical flow 50 in the vehicle-mounted camera deposit | attachment detection apparatus 9. FIG. FIG. 5 is a diagram showing a search range 40 at a measurement point 20 of an optical flow 50. It is a figure which shows the measuring method of the optical flow 50. FIG. It is a figure which shows the scene where the vehicle is passing the rhombus road marking 60. FIG. It is a figure which shows the estimated optical flow 70. FIG. It is a figure which shows the optical flow 50 measured in the case where there exists an adhesion thing, and when there is no attachment. The flowchart which shows the operation example in the adhesion detection apparatus 9 for vehicle-mounted cameras.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Imaging device, 2 ... A / D conversion circuit, 3 ... Image processing device, 4 ... Memory, 5 ... Arithmetic device, 6 ... Car navigation system, 7 ... Vehicle information, 8 ... Alarm device, 9 ... For vehicle-mounted camera Kimono detection device, 10 ... vehicle, 11 ... imaging range, 13 ... optical system optical component, 20 ... optical flow measurement point, 21 ... optical flow measurement range, 40 ... optical flow search range at optical flow measurement point, 50 ... measured 60 ... road marking, 61 ... white line, 70 ... estimated optical flow, 80 ... deposit.

Claims (19)

In an in-vehicle camera that includes an optical system, an image sensor, and an arithmetic device that processes an image captured by the image sensor,
An optical flow measuring means for calculating a moving amount and a moving direction of a predetermined extract from a plurality of images taken at different timings by the imaging device;
Optical flow estimation means for estimating the amount and direction of movement of the extract based on vehicle speed information,
Based on the movement amount and movement direction of the extract measured by the optical flow measurement means and the movement amount and movement direction of the extract estimated by the optical flow estimation means, the optical system of the imaging means An in-vehicle camera that detects attached material. The in-vehicle camera according to claim 1, wherein the optical flow estimation unit estimates a moving amount and a moving direction of the extract using yaw rate sensor information in addition to the vehicle speed information. In claim 1,
The in-vehicle camera characterized in that the optical flow estimation means estimates a moving amount and a moving direction of the extract using steering angle sensor information in addition to the vehicle speed information. In any one of Claims 1 thru | or 3,
The vehicle is obtained by acquiring the vehicle position and map information by GPS via a communication path, and detecting the attached object when it is predicted that there is a predetermined target within the imaging range of the vehicle-mounted camera. camera. In claim 4,
A vehicle-mounted camera using paint on a road as the predetermined target. In any one of Claims 1 thru | or 5,
An in-vehicle camera that communicates with a car navigation system, detects an adhering substance based on communication information from the car navigation system, and changes a processing sequence. In claim 6,
A vehicle-mounted camera that detects the adhering matter when information indicating that there is a road marking paint within the imaging range of the vehicle-mounted camera is received from the car navigation system. In any one of Claims 1 thru | or 7,
An on-vehicle camera characterized by narrowing an optical flow search area in the optical flow measurement means when the load on the arithmetic unit is high. In any one of Claims 1 thru | or 7,
An in-vehicle camera, wherein the number of optical flow measurement points in the optical flow measurement means is reduced when the load on the arithmetic unit is high. In any one of Claims 1 thru | or 7,
An in-vehicle camera, wherein an optical flow search area or an optical flow measurement point in the optical flow measuring means is changed based on vehicle speed information and yaw rate sensor information. In any one of Claims 1 thru | or 7,
An in-vehicle camera characterized in that an optical flow search area or an optical flow measurement point in the optical flow measuring means is changed based on vehicle speed information and rudder angle sensor information. In any one of Claims 1 thru | or 7,
An in-vehicle camera characterized in that, based on GPS information, an optical flow search area or the number of optical flow measurement points in the optical flow measurement means is changed. In any one of Claims 1 thru | or 12,
A vehicle-mounted camera, wherein the frequency of executing the attached matter detection process is changed based on vehicle speed information. In any one of Claims 1 thru | or 12,
An in-vehicle camera that determines whether or not to execute the process of detecting the attached matter based on information on a transmission provided in the vehicle. In any one of Claims 1 thru | or 14,
An in-vehicle camera that determines whether or not to execute the process of detecting the attached matter based on a load factor of the arithmetic device. In any one of Claims 1 thru | or 15,
An in-vehicle camera that determines whether or not to execute the process of detecting the attached matter based on wiper information. In any one of Claims 1 to 16,
An in-vehicle camera that outputs alarm information when it is recognized that there is a deposit in the optical system. In any one of Claims 1 thru | or 17,
An in-vehicle camera that controls an exposure time of the image sensor based on a luminance distribution of a captured image. The in-vehicle camera according to claim 18, wherein a gain for the electric charge accumulated in the image sensor is controlled based on a luminance distribution of a captured image.

Images