Method for judging contamination of vehicle-mounted camera, system for judging contamination, and method for cleaning

Abstract

This invention discloses a contamination judgment method, a contamination judgment system, and a cleaning method for a vehicle-mounted camera. The contamination judgment method includes the following steps: allowing a vehicle-mounted camera to capture a poster containing scene information through contaminated specimens of different gray levels to obtain a series of specimen images; converting the original image data of the series of specimen images into HSV format image data; extracting the mean and variance of the HSV color space of the HSV format image data of the series of specimen images as features and inputting them into a support vector machine algorithm model to train a classification algorithm model; inputting the image data of a detection image captured by the vehicle-mounted camera as input features into the classification algorithm model to allow the classification algorithm model to classify the detection image captured by the vehicle-mounted camera; and determining whether the lens surface of the vehicle-mounted camera is contaminated based on the classification result of the classification algorithm model.

Description

Technical Field

[0001] This invention relates to the automotive field, and in particular to a method for determining contamination of an in-vehicle camera, a contamination determination system, and a cleaning method. Background Technology

[0002] Vehicle-mounted cameras, as key sensors for collecting environmental information and sensing environmental conditions, are important components of vehicles and have a wide range of applications. The lens surface (i.e., the optical window) of a vehicle-mounted camera is exposed to the vehicle's environment, allowing ambient light to enter and capture images of the surrounding environment. Due to the complex environment in which vehicles operate, dust, water vapor, and other contaminants can easily adhere to the lens surface of the vehicle-mounted camera, preventing it from capturing clear images. Therefore, how to determine whether dust, water vapor, or other contaminants adhere to the lens surface of a vehicle-mounted camera, thus preventing clear imaging, is the technical problem that the inventors of this invention have been committed to solving. Summary of the Invention

[0003] One object of the present invention is to provide a method for judging contamination of a vehicle-mounted camera, a contamination judgment system, and a cleaning method, wherein the contamination judgment method classifies the detection images captured by the vehicle-mounted camera based on a classification algorithm model, and can determine whether the lens surface of the vehicle-mounted camera is contaminated based on the classification results.

[0004] One object of the present invention is to provide a pollution judgment method, a pollution judgment system and a cleaning method for a vehicle-mounted camera, wherein the pollution judgment method can provide a large number of sample images to train the classification algorithm model, so as to efficiently and accurately classify the detection images captured by the vehicle-mounted camera in the subsequent process.

[0005] One object of the present invention is to provide a pollution judgment method, a pollution judgment system and a cleaning method for a vehicle-mounted camera, wherein the pollution judgment method can control the vehicle-mounted camera to take pictures in different scenarios and / or through different polluted samples to obtain a large number of sample images.

[0006] One object of the present invention is to provide a pollution judgment method, a pollution judgment system and a cleaning method for a vehicle-mounted camera, wherein the pollution judgment method can collect a large number of specimen images in an indoor environment for use in training the classification algorithm model, thereby reducing the workload of collecting specimen images outdoors and eliminating the safety hazards when collecting specimen images outdoors.

[0007] One object of the present invention is to provide a method for determining contamination of a vehicle-mounted camera, a contamination determination system, and a cleaning method, wherein when the lens surface of the vehicle-mounted camera is contaminated, the cleaning method can clean the vehicle-mounted camera by spraying fluid onto the lens surface of the vehicle-mounted camera through a nozzle.

[0008] According to one aspect of the present invention, the present invention provides a method for determining contamination of an in-vehicle camera, wherein the contamination determination method includes the following steps:

[0009] (a) In an indoor scene platform environment, a vehicle-mounted camera is allowed to capture a poster containing scene information through contaminated specimens of different gray levels to obtain a series of specimen images, wherein step (a) further includes the step of:

[0010] (a.1) Construct a specimen image acquisition platform indoors;

[0011] (a.2) The poster is placed at one end of the specimen image acquisition stand, in the specimen image acquisition...

[0012] The vehicle-mounted camera is arranged at the other end of the platform;

[0013] (a.3) The contaminated specimen is placed between the vehicle-mounted camera and the poster to allow the vehicle-mounted camera to...

[0014] The camera takes a picture of the poster through the contaminated specimen, wherein the contaminated specimen is replaceable;

[0015] (b) Convert the raw image data of the series of specimen images into image data in HSV format;

[0016] (c) The mean and variance of the HSV color space of the HSV format image data of a series of specimen images are used as features and input into the support vector machine algorithm model to train a classification algorithm model.

[0017] (d) Input the image data of a detection image captured by the vehicle-mounted camera as a feature into the classification algorithm model, so as to allow the classification algorithm model to classify the detection image captured by the vehicle-mounted camera; (e) Determine whether the lens surface of the vehicle-mounted camera is contaminated based on the classification result of the classification algorithm model.

[0018] According to one embodiment of the present invention, in step (a), the vehicle-mounted camera takes a picture of the poster through the contaminated specimen under natural light conditions.

[0019] According to one embodiment of the present invention, after step (a.2), step (a) further includes the step of: (a.4) calibrating the field of view of the vehicle-mounted camera.

[0020] According to one embodiment of the present invention, the scene information in the poster is pedestrians crossing the road, pedestrians walking along the road, two-wheeled vehicles crossing the road, or two-wheeled vehicles riding along the road.

[0021] According to another aspect of the present invention, the present invention further provides a method for cleaning a vehicle-mounted camera, wherein the cleaning method includes the following steps:

[0022] (A) Image data of a detection image captured by a vehicle-mounted camera is used as a feature input to a classification algorithm model, so that the classification algorithm model can classify the detection image captured by the vehicle-mounted camera;

[0023] (B) Based on the classification results of the classification algorithm model, determine whether the lens surface of the vehicle camera is contaminated;

[0024] (C) When it is determined that the lens surface of the vehicle camera is contaminated, a nozzle is allowed to spray fluid onto the lens surface of the vehicle camera to clean the vehicle camera; wherein step (C) further includes step (C.1) increasing the pressure of the cleaning fluid in a tube passage of a telescopic tube of the nozzle to allow the cleaning fluid to push the telescopic tube to move relative to a sleeve of the nozzle and expose a nozzle body.

[0025] (C.2) Increase the pressure of the cleaning fluid in the tube channel of the telescopic tube to allow the cleaning fluid to push a sealing element of the nozzle to deform and form a liquid flow channel between the sealing element and the telescopic tube, wherein the liquid flow channel connects a central outlet of the telescopic tube and an annular outlet groove.

[0026] (C.3) The cleaning fluid in the tube channel of the telescopic tube is allowed to sequentially pass through the central outlet of the telescopic tube, the annular outlet groove and a peripheral outlet channel, the outlet channel of a nozzle bracket of the nozzle and the nozzle body, and be sprayed out of the nozzle and onto the lens surface of the vehicle camera to clean the vehicle camera.

[0027] According to another aspect of the present invention, the present invention further provides a pollution detection system for a vehicle-mounted camera, comprising:

[0028] A format conversion module is used to convert the raw image data of a series of specimen images captured by a vehicle-mounted camera into HSV format image data;

[0029] A feature extraction module is used to extract the mean and variance of the HSV color space of the HSV format image data of a series of specimen images as features;

[0030] A training module is used to input the features of a series of specimen images into a support vector machine algorithm model to train a classification algorithm model;

[0031] A classification module is used to input the image data of a detection image captured by the vehicle-mounted camera as a feature into the classification algorithm model, so that the classification algorithm model can classify the detection image captured by the vehicle-mounted camera.

[0032] A judgment module is used to determine whether the lens surface of the vehicle camera is contaminated based on the classification result of the classification algorithm model.

[0033] According to one embodiment of the present invention, the contamination judgment system further includes a construction module for constructing a specimen image acquisition platform indoors, wherein a poster containing scene information and the vehicle-mounted camera can be respectively arranged at opposite ends of the specimen image acquisition platform, and a contaminated specimen is placed between the vehicle-mounted camera and the poster, so that the vehicle-mounted camera can capture the poster through the contaminated specimen to obtain the specimen image.

[0034] According to one embodiment of the present invention, the contaminated specimen is replaceable.

[0035] According to one embodiment of the present invention, the pollution determination system further includes a calibration module for use in a

[0036] When a checkerboard pattern is set at the end of the specimen image acquisition platform, the field of view of the vehicle-mounted camera is calibrated.

[0037] According to one embodiment of the present invention, after the field of view of the vehicle-mounted camera is calibrated, the poster is used to replace the chessboard grid.

[0038] According to one embodiment of the present invention, the scene information in the poster is pedestrians crossing the road, pedestrians walking along the road, two-wheeled vehicles crossing the road, or two-wheeled vehicles riding along the road.

[0039] The beneficial effects of this invention are:

[0040] First, the pollution judgment method reduces the workload of collecting specimen images outdoors and eliminates safety hazards when collecting specimen images outdoors by controlling the vehicle-mounted camera to capture the specimen image of the poster through the polluted specimen on the indoor scene platform, and by making the polluted specimen replaceable.

[0041] Secondly, the pollution judgment method, by placing the vehicle-mounted camera and the poster on the specimen image acquisition platform respectively, and by interchangeably placing the polluted specimens of different gray levels between the vehicle-mounted camera and the poster, can acquire a large number of specimen images in an indoor environment, which is beneficial for training the classification algorithm model and thus ensuring the accuracy of subsequent judgment results.

[0042] Furthermore, the specimen image acquisition platform is an open platform to allow natural light to enter its contents. Thus, compared to artificial light simulation, the specimen images acquired by the pollution judgment method of the present invention have a more realistic simulation effect, thereby ensuring the accuracy of subsequent judgment results.

[0043] Fourth, before the pollution judgment method controls the vehicle-mounted camera to capture the sample image of the poster through the pollution specimen, it can calibrate the field of view of the vehicle-mounted camera to improve the image quality of the sample image.

[0044] Fifth, when the contamination judgment method determines that the lens surface of the vehicle camera is contaminated, the cleaning method can spray fluid onto the lens surface of the vehicle camera through the nozzle disposed adjacent to the vehicle camera to clean the contaminants adhering to the lens surface of the vehicle camera, thereby improving the imaging quality of the vehicle camera and benefiting driving safety.

[0045] Other beneficial effects of the present invention will be disclosed and described in detail below. Attached Figure Description

[0046] Figure 1 The flowchart of a pollution determination method according to the present invention is shown.

[0047] Figure 2 The three-dimensional state of a sample image acquisition stand according to the present invention is shown.

[0048] Figure 3 The diagram illustrates the state of calibrating the field of view of an in-vehicle camera according to the present invention.

[0049] Figure 4 The image shows a poster being photographed through a contaminated specimen by the vehicle-mounted camera according to the present invention.

[0050] Figure 5A The state of the first specimen image captured by the vehicle-mounted camera according to the present invention is shown.

[0051] Figure 5B The state of the second specimen image captured by the vehicle-mounted camera according to the present invention is shown.

[0052] Figure 5C The state of the third specimen image captured by the vehicle-mounted camera according to the present invention is shown.

[0053] Figure 5D The state of the fourth specimen image captured by the vehicle-mounted camera according to the present invention is shown.

[0054] Figure 6 A block diagram of a pollution detection system according to the present invention is shown.

[0055] Figure 7 A block diagram of an electronic device according to the present invention is shown.

[0056] Figure 8 The flow chart of a cleaning method according to the present invention is shown.

[0057] Figure 9 The three-dimensional state of a nozzle according to the present invention is shown.

[0058] Figure 10 The disassembled state of the nozzle according to the present invention is shown.

[0059] Figure 11 The nozzle according to the present invention is shown in cross-section.

[0060] Figure 12A A cross-sectional schematic diagram of a portion of the nozzle according to the present invention in a non-operating state is shown.

[0061] Figure 12B A cross-sectional schematic diagram of a portion of the nozzle according to the present invention in the working state is shown.

[0062] Explanation of reference numerals in the attached figures:

[0063] 100. Vehicle-mounted camera; 200. Contaminated specimen; 300. Poster; 400. Specimen image acquisition platform; 500. Checkerboard pattern; 501. Calibration pattern; 600. Platform; 10. Format conversion module; 20. Feature extraction module; 30. Training module; 40. Classification module; 50. Judgment module; 60. Construction module; 70. Calibration module; 1000. Processor; 2000. Memory; 3000. Input device; 4000. Output device; 80. Nozzle; 81. Sleeve; 811. Telescopic space; 812. Liquid inlet channel; 82. Nozzle support; 821. Deformation space; 822. Vent; 82 3. Liquid outlet channel; 824. Installation space; 83. Nozzle body; 84. Telescopic tube; 841. Tube body channel; 842. Liquid inlet; 843. Central liquid outlet; 844. Annular liquid outlet groove; 845. Peripheral liquid outlet through hole; 85. Sealing element; 86. Reset seat; 861. Seat space; 87. First reset element; 88. Assembly seat; 881. Assembly ring; 8811. Ring channel; 882. First assembly tube; 8821. First assembly space; 883. Second assembly tube; 8831. Second assembly space; 829. Second reset element; 810. Sealing seat; 801. Liquid channel; 802. Liquid holding space. Detailed Implementation

[0064] Before detailing any embodiment of the invention, it should be understood that the invention is not limited in its application to the details of the construction and arrangement of the components set forth in the following description or illustrated in the following figures. The invention can have other embodiments and can be practiced or carried out in various ways. Furthermore, it should be understood that the wording and terminology used herein are for descriptive purposes and should not be considered limiting. The use of “comprising,” “including,” or “having,” and variations thereof is intended to cover the items set forth below and their equivalents, as well as any additional items. Unless otherwise specified or limited, the terms “installation,” “connection,” “support,” and “linkage,” and variations thereof are used broadly and cover both direct and indirect installation, connection, support, and linking. Moreover, “connection” and “linkage” are not limited to physical or mechanical connections or links.

[0065] Furthermore, firstly, in the disclosure of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the above terms should not be construed as limiting this invention. Secondly, the term "a" should be understood as "at least one" or "one or more," that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple. The term "a" should not be construed as a limitation on the quantity.

[0066] Refer to the accompanying drawings of the specification of this invention. Figure 1 A contamination determination method according to a preferred embodiment of the present invention will be disclosed and described in the following description. The contamination determination method is used to determine whether a vehicle-mounted camera 100 is contaminated, and the contamination determination method includes the following steps:

[0067] S110, in an indoor scene platform environment, the vehicle-mounted camera 100 is allowed to capture a poster 300 containing scene information through contaminated specimens 200 of different gray levels to obtain a series of specimen images;

[0068] S120, convert the raw image data of the series of specimen images into image data in HSV format;

[0069] S130, the mean and variance of the HSV color space of the HSV format image data of a series of specimen images are extracted as features and input into a support vector machine algorithm model to train a classification algorithm model.

[0070] S140, the image data of a detection image captured by the vehicle-mounted camera 100 is input as a feature into the classification algorithm model, so that the classification algorithm model can classify the detection image captured by the vehicle-mounted camera 100; and;

[0071] S150, based on the classification result of the classification algorithm model, determine whether the lens surface of the vehicle camera 100 is contaminated.

[0072] In other words, the contamination determination method of the present invention can capture a large number of specimen images using the vehicle-mounted camera 100, and use these specimen images to train the classification algorithm model. After the image data of the detection images captured by the vehicle-mounted camera 100 is input into the classification algorithm model, the classification algorithm model can efficiently and accurately classify the detection images captured by the vehicle-mounted camera 100, so as to determine whether the lens surface of the vehicle-mounted camera 100 is contaminated based on the classification result of the classification algorithm model. Furthermore, based on the classification result of the classification algorithm model, the contamination determination method can determine the degree of contamination on the lens surface of the vehicle-mounted camera 100.

[0073] When training the classification algorithm model, the pollution judgment method of the present invention requires a large number of specimen images. In order to make the specimen images have strong generalization ability and robust performance, the pollution judgment method of the present invention needs to control the vehicle-mounted camera 100 to take pictures of the poster 300 under different scenes and / or through the polluted specimens 200 with different gray levels and / or through the polluted specimens 200 to take pictures of the poster 300 containing different scene information, so as to obtain a large number of specimen images.

[0074] To reduce the workload of collecting specimen images outdoors and eliminate safety hazards during outdoor specimen image collection, the contamination judgment method of the present invention controls the vehicle-mounted camera 100 to capture images of the poster 300 through the contaminated specimen 200 in an indoor scene platform environment, such as... Figures 2 to 5D As shown. In other words, in step S110, the vehicle-mounted camera 100 captures an image of the poster 300 through the contaminated specimen 200 in the indoor scene platform environment to obtain the specimen image.

[0075] Specifically, step S110 further includes the following steps: S1101, setting up a specimen image acquisition platform 400 indoors; S1102, arranging the poster 300 at one end of the specimen image acquisition platform 400, and arranging the vehicle-mounted camera 100 at the other end of the specimen image acquisition platform 400; S1103, placing the contaminated specimen 200 between the vehicle-mounted camera 100 and the poster 300, so as to allow the vehicle-mounted camera 100 to photograph the poster 300 through the contaminated specimen 200.

[0076] It is worth noting that the manner in which the vehicle-mounted camera 100 is positioned at the end of the specimen image acquisition platform 400 is not limited in the contamination determination method of the present invention. For example, the vehicle-mounted camera 100 can be positioned at the end of the specimen image acquisition platform 400 via a platform 600. Preferably, the platform 600 is configured to allow adjustment of the position and angle of the vehicle-mounted camera 100.

[0077] It is worth noting that the method of placing the contaminated specimen 200 between the vehicle-mounted camera 100 and the poster 300 is not limited in the contamination determination method of the present invention. For example, the contaminated specimen 200 may include a light-transmitting substrate and contaminants attached to the light-transmitting substrate. The light-transmitting substrate can be detachably mounted on the specimen image acquisition stand 400 to place the contaminated specimen 200 between the vehicle-mounted camera 100 and the poster 300. By replacing the contaminated specimen 200 with different gray levels, the vehicle-mounted camera 100 can obtain different specimen images when photographing the poster 300 through the contaminated specimen 200 with different gray levels.

[0078] In some embodiments of the contamination determination method of the present invention, the poster 300 may be a physical poster, which is detachably mounted on the end of the specimen image acquisition platform 400. By replacing the contaminated specimen 200 with different grayscale levels and the poster 300 with different scene information, the vehicle-mounted camera 100 can capture a large number of specimen images, thereby improving the generalization ability and robustness of the sample images. In other embodiments of the contamination determination method of the present invention, the poster 300 may be a virtual poster, which is projected onto the end of the specimen image acquisition platform 400. By replacing the contaminated specimen 200 with different grayscale levels and projecting the poster 300 with different scene information, the vehicle-mounted camera 100 can capture a large number of specimen images, thereby improving the generalization ability and robustness of the sample images, increasing the specimen image acquisition speed, and thus improving the efficiency of training the classification algorithm model.

[0079] Preferably, the scene information in the poster 300 is pedestrians crossing the road, pedestrians walking along the road, two-wheeled vehicles crossing the road, or two-wheeled vehicles riding along the road, to simulate a real scene. Preferably, to further reproduce a real scene, photos containing road zebra crossings can be posted on the perimeter walls of the specimen image acquisition platform 400.

[0080] After step S1102, step S110 further includes the step of calibrating the field of view of the vehicle-mounted camera 100.

[0081] Specifically, see the attached document. Figure 2 The specimen image acquisition platform 400 is constructed using a 160cm × 64cm × 64cm frame, and the visible area of ​​the poster 300 is 37cm × 20.5cm. In other words, the field of view of the vehicle-mounted camera 100 on the poster 300 is 37cm × 20.5cm. Those skilled in the art will understand that the 37cm × 20.5cm field of view of the vehicle-mounted camera 100 on the poster 300 does not mean that the size of the poster 300 is only 37cm × 20.5cm; that is, the size of the poster 300 can be larger than 37cm × 20.5cm. Preferably, the size of the poster 300 is larger than 37cm × 20.5cm; for example, the size of the poster 300 can reach 64cm × 64cm, to allow the poster 300 to be adapted to different types of vehicle-mounted cameras 100.

[0082] To ensure that the field of view of the vehicle-mounted camera 100 on the poster is 37cm × 20.5cm, the pollution judgment method of the present invention uses a checkerboard calibration method to calibrate the field of view of the vehicle-mounted camera 100 on the poster 300, wherein the checkerboard 500 used for calibration is as follows: Figure 3 As shown, the checkerboard 500 has multiple circular calibration patterns 501 arranged in four rows. Each calibration pattern 501 has a diameter of 4 cm. The distance between the centers of two adjacent calibration patterns 501 in the same row is 11 cm, and the distance between the centers of calibration patterns 501 in adjacent rows is 5.5 cm. When calibrating the field of view of the vehicle-mounted camera 100 using the checkerboard calibration method, the focal length of the vehicle-mounted camera 100 is adjusted to align its field of view with the position of the circular calibration patterns 501 in the checkerboard 500, thereby achieving the calibration of the vehicle-mounted camera 100's field of view.

[0083] It is worth mentioning that, in some embodiments of the pollution judgment method of the present invention, the checkerboard 500 is a physical checkerboard that is detachably installed at one end of the specimen image acquisition platform 400. Thus, after calibrating the field of view of the vehicle-mounted camera 100 using the checkerboard 500, the checkerboard 500 can be detached from the end of the specimen image acquisition platform 400, and the poster 300 can be placed at the end of the specimen image acquisition platform 400. Furthermore, during the replacement of the checkerboard 500 by the poster 300, the positions of the vehicle-mounted camera 100 and the specimen image acquisition platform 400 remain unchanged. Therefore, after the poster 300 replaces the checkerboard 500, the field of view of the vehicle-mounted camera 100 on the poster 300 is guaranteed to be the calibrated field of view. In some other embodiments of the pollution judgment method of the present invention, the checkerboard 500 is a virtual checkerboard that is projected onto the end of the specimen image acquisition platform 400. After the field of view of the vehicle-mounted camera 100 is calibrated using the checkerboard 500, the poster 300 can be projected onto the end of the specimen image acquisition platform 400. In this way, it can be ensured that the field of view of the vehicle-mounted camera 100 on the poster 300 is the calibrated field of view, while improving the efficiency of replacing the checkerboard 500 with the poster 300.

[0084] In some embodiments of the contamination determination method of the present invention, the peripheral walls of the specimen image acquisition platform 400 are open to allow natural light to enter the space defined by the specimen image acquisition platform 400, thereby allowing the vehicle-mounted camera 100 to capture the poster 300 in natural light through the contaminated specimen 200 to obtain the specimen image.

[0085] Reference Appendix Figures 4 to 5D After calibrating the field of view of the vehicle-mounted camera 100, the checkerboard 500 is replaced with the poster 300. At this time, the field of view of the vehicle-mounted camera 100 on the poster 300 is the calibrated field of view. The vehicle-mounted camera 100 can take pictures of the poster 300 through the polluted specimen 200 under natural conditions to obtain the specimen image. It can be understood that, on the one hand, by sequentially replacing the polluted specimen 200 with different grayscale levels and allowing the vehicle-mounted camera 100 to take pictures of the poster 300 through the polluted specimen 200, a series of specimen images can be obtained. On the other hand, by sequentially replacing the poster 300 with different scene information and allowing the vehicle-mounted camera 100 to take pictures of the poster 300 through the polluted specimen 200, a series of specimen images can be obtained.

[0086] After obtaining a series of specimen images, the contamination determination method converts the original image data of the series of specimen images into HSV format, wherein in the HSV color space model:

[0087] Hue (H): Hue is measured in degrees, ranging from 0° to 360°. Starting with red and counting counter-clockwise, red is 0°, green is 120°, and blue is 240°. Their complementary colors are yellow (60°), cyan (180°), and magenta (300°).

[0088] Saturation (S): Saturation (S) indicates how close a color is to a spectral color. Any color can be considered the result of mixing a spectral color with white. The greater the proportion of the spectral color, the closer the color is to the spectral color, and the higher the saturation. High saturation results in a deep and vibrant color. Saturation reaches its maximum when the white light component in the spectral color is zero. The saturation range is 0%-100%; the higher the value, the more saturated the color.

[0089] Brightness (V): Brightness indicates the degree of lightness of a color. For light source color, the brightness value is related to the luminance of the light source; for object color, this value is related to the object's transmittance or reflectance. The brightness value ranges from 0% (black) to 100% (white).

[0090] Statistical analysis was performed on various information items in the original image data and HSV format image data of a series of specimen images captured by the vehicle-mounted camera 100. A comparison of the two revealed that as the degree of contamination in the specimen image increased, the mean value of the image saturation S decreased. Simultaneously, the variance of each HSV data item decreased, indicating that the data contrast was decreasing and edge features were becoming less distinct. It can be understood that the degree of contamination in the specimen image corresponds to the grayscale level of the contaminated specimen 200. The higher the grayscale level of the contaminated specimen 200, the more severe the contamination of the sample image obtained by the vehicle-mounted camera 100 after capturing the poster 300 through the contaminated specimen 200.

[0091] In low-light conditions at night, the insufficient light intake of the vehicle-mounted camera 100 prevents it from capturing more information reflecting the environmental conditions, resulting in lower image quality compared to daytime conditions. A comparison of the statistical information of the image data of the sample image of the poster 300 directly captured by the vehicle-mounted camera 100 (i.e., the characteristic values ​​of the sample image data in the HSV color space model) with the statistical information of the image data of the sample image of the poster 300 captured by the vehicle-mounted camera 100 through the pollution sample 200 reveals that, in low-light conditions at night, the mean saturation S of the image data of the sample image of the poster 300 captured by the vehicle-mounted camera 100 through the pollution sample 200 is lower and also lower than the mean saturation S of the image data of the sample image of the poster 300 directly captured by the vehicle-mounted camera 100. Therefore, compared to the normal specimen image (i.e., the sample image of the poster 300 directly captured by the vehicle-mounted camera 100), the contaminated sample image (i.e., the sample image of the poster 300 captured by the vehicle-mounted camera 100 through the contaminated specimen 200) has relatively obvious features in each feature value in the HSV color space model. Therefore, the mean and variance of the HSV color space of the image data can be used as features for extraction and input into the support vector machine algorithm model to train the classification algorithm model.

[0092] Since the contaminated sample images have relatively obvious features in each feature value of the HSV color space model, the contamination judgment method can input a large number of features (mean and variance) of the contaminated sample images in the HSV color space into the support vector machine algorithm model to train the classification algorithm model.

[0093] When it is necessary to determine whether the lens surface of the vehicle-mounted camera 100 is contaminated, firstly, the statistical information of the image data of the detection image captured by the vehicle-mounted camera 100 (i.e., the feature values ​​of the image data of the detection image in the HSV color space model) can be used as features input into the classification algorithm model. Secondly, the classification algorithm model classifies the detection image captured by the vehicle-mounted camera 100. Finally, based on the classification result of the classification algorithm model, it can be determined whether the lens surface of the vehicle-mounted camera 100 is contaminated.

[0094] According to another aspect of the invention, reference is made to the appendix. Figure 6The present invention further provides a contamination judgment system for judging whether the lens surface of the vehicle camera 100 is contaminated, wherein the contamination judgment system includes a format conversion module 10, a feature extraction module 20, a training module 30, a classification module 40 and a judgment module 50.

[0095] Specifically, the format conversion module 10 is used to convert the original image data of a series of specimen images captured by the vehicle-mounted camera 100 into HSV format image data. The feature extraction module 20 is used to extract the mean and variance of the HSV color space of the HSV format image data of the series of specimen images as features. The training module 30 is used to input the features of the series of specimen images into the support vector machine algorithm model to train the classification algorithm model. The classification module 40 is used to input the image data of the detection images captured by the vehicle-mounted camera 100 as features into the classification algorithm model so that the classification algorithm model can classify the detection images captured by the vehicle-mounted camera 100. The judgment module 50 is used to determine whether the lens surface of the vehicle-mounted camera 100 is contaminated based on the classification result of the classification algorithm model.

[0096] Furthermore, the contamination judgment system includes a setup module 60, which is used to set up the specimen image acquisition platform 400 indoors. The poster 300 and the vehicle-mounted camera 100 can be respectively arranged at opposite ends of the specimen image acquisition platform 400. The contaminated specimen 200 is placed between the vehicle-mounted camera 100 and the poster 300, so that the vehicle-mounted camera 100 can capture the poster 300 through the contaminated specimen 200 to obtain the specimen image. Subsequently, the format conversion module 10 can convert the original image data of the specimen image into HSV format image data.

[0097] Preferably, the contaminated specimen 200 is replaceable. For example, the contaminated specimen 200 is detachably mounted on the specimen image acquisition platform 400 to hold the contaminated specimen 200 between the vehicle-mounted camera 100 and the poster 300, so that the vehicle-mounted camera 100 can capture the poster 300 through the contaminated specimen 200 to obtain the sample image. By replacing the contaminated specimen 200 with specimens of different gray levels, a large number of specimen images can be acquired for training the classification algorithm model. Furthermore, in this way, the sample images exhibit strong generalization ability and robustness.

[0098] Preferably, the poster 300 is replaceable. For example, the poster 300 is detachably mounted on the sample image acquisition stand 400, or the poster 300 is replaceably projected onto the sample image acquisition stand 400. By replacing the poster 300 with different scene information, a large number of sample images can be acquired for training the classification algorithm model. Furthermore, in this way, the sample images exhibit strong generalization ability and robustness.

[0099] In some embodiments of the contamination judgment system of the present invention, the sample image acquisition platform 400 constructed by the construction module 60 has a frame structure with dimensions of 160cm × 64cm × 64cm to allow natural light to enter the interior of the sample image acquisition platform 400, thereby allowing the vehicle-mounted camera 100 to photograph the poster 300 through the contaminated specimen 200 under natural light conditions. The poster 300 has a size greater than 37cm × 20.5cm. To ensure that the field of view of the vehicle-mounted camera 100 on the poster 300 is 37cm × 20.5cm, the contamination judgment system of the present invention further includes a calibration module 70, which is used to calibrate the field of view of the vehicle-mounted camera 100 when the checkerboard 500 is set at the end of the sample image acquisition platform 400.

[0100] Specifically, the checkerboard 500 has multiple circular calibration patterns 501 arranged in four rows. Each calibration pattern 501 has a diameter of 4 cm. The distance between the centers of two adjacent calibration patterns 501 in the same row is 11 cm, and the distance between the centers of calibration patterns 501 in adjacent rows is 5.5 cm. When calibrating the field of view of the vehicle-mounted camera 100 using the checkerboard calibration method, the focal length of the vehicle-mounted camera 100 is adjusted to align its field of view with the position of the circular calibration patterns 501 in the checkerboard 500, thereby achieving the calibration of the vehicle-mounted camera 100's field of view. After the calibration module 70 calibrates the field of view of the vehicle-mounted camera 100, the poster 300 is used to replace the checkerboard 500 so that the vehicle-mounted camera 100 can subsequently capture the poster 300 under natural light conditions through the contaminated specimen 200 to obtain the sample image.

[0101] Here, those skilled in the art will understand that the specific functions and operations of each module of the pollution detection system have been described in the appendix above. Figures 1 to 5D The pollution determination method described herein has been described in detail, and therefore, its repeated description will be omitted.

[0102] It is worth mentioning that the pollution detection system of the present invention can be implemented in various terminal devices, such as computers or devices with computing capabilities. In some embodiments, the pollution detection system can be integrated into the terminal device as a software module and / or a hardware module. For example, the pollution detection system can be a software module in the operating system of the terminal device, or it can be an application developed for the terminal device. Of course, the pollution detection system can also be one of many hardware modules of the terminal device.

[0103] According to another aspect of the invention, reference is made to the appendix. Figure 7 The present invention further provides an electronic device, wherein the electronic device includes at least one processor 1000 and at least one memory 2000.

[0104] The processor 1000 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0105] The memory 2000 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory, wherein the volatile memory may include random access memory (RAM) and / or advanced cache memory, and wherein the non-volatile memory may include read-only memory (ROM), hard disk, flash memory, etc.

[0106] One or more computer program instructions may be stored in the memory 2000, and the processor 1000 is capable of reading the computer program instructions from the memory 2000 and executing the computer program instructions to implement the pollution determination method and pollution determination system disclosed above.

[0107] Preferably, the electronic device further includes an input device 3000 and an output device 4000, which can be interconnected via a bus system and / or other forms of connection.

[0108] According to another aspect of the invention, reference is made to the appendix. Figure 8 The present invention further provides a cleaning method for cleaning the lens surface of the vehicle-mounted camera 100, wherein the cleaning method includes the following steps:

[0109] Step 1001: The image data of the detection image captured by the vehicle-mounted camera 100 is used as a feature input into the classification algorithm model, so as to allow the classification algorithm model to classify the detection image captured by the vehicle-mounted camera 100;

[0110] Step 1002: Based on the classification result of the classification algorithm model, determine whether the lens surface of the vehicle-mounted camera 100 is contaminated; and

[0111] Step 1003: When it is determined that the lens surface of the vehicle camera 100 is contaminated, a nozzle 80 is allowed to spray fluid onto the lens surface of the vehicle camera 100 to clean the vehicle camera 100.

[0112] Preferably, the nozzle 80 and the vehicle-mounted camera 100 can be arranged adjacent to each other, so that when the lens surface of the vehicle-mounted camera 100 is contaminated, the nozzle 80 can clean the vehicle-mounted camera 100 by spraying fluid onto the lens surface of the vehicle-mounted camera 100 to remove contaminants adhering to the lens surface of the vehicle-mounted camera 100, so that the vehicle-mounted camera 100 can form a clear image.

[0113] Preferably, when the lens surface of the vehicle camera 100 is contaminated, the nozzle 80 can clean the vehicle camera 100 by spraying cleaning fluid onto the lens surface of the vehicle camera 100 to remove contaminants adhering to the lens surface of the vehicle camera 100.

[0114] Reference Appendix Figures 9 to 12B The nozzle 80 includes a sleeve 81, a nozzle support 82, a nozzle body 83, a telescopic tube 84, a deformable sealing element 85, a reset seat 86, and a first reset element 87.

[0115] The sleeve 81 has a telescopic space 811. The nozzle support 82 has a deformation space 821, a vent 822 and at least one liquid outlet channel 823. The vent 822 is connected to the deformation space 821. The nozzle body 83 is mounted on the nozzle support 82 and is connected to the liquid outlet channel 823 of the nozzle support 82. The telescopic tube 84 is telescopically installed in the telescopic space 811 of the sleeve 81. The telescopic tube 84 has a tube body channel 841, a liquid inlet 842, a central liquid outlet 843, an annular liquid outlet groove 844, and at least one peripheral liquid outlet through hole 845. The liquid inlet 842 and the central liquid outlet 843 are respectively connected to the tube body channel 841 at opposite ends of the telescopic tube 84. The annular liquid outlet groove 844 surrounds the central liquid outlet 843, and the peripheral liquid outlet through hole 845 is connected to the annular liquid outlet groove 844. The nozzle bracket 82 is installed on the telescopic tube 84, and the peripheral liquid outlet through hole 845 of the telescopic tube 84 is connected to the liquid outlet channel 823 of the nozzle bracket 82. The sealing element 85 is deformably held between the nozzle support 82 and the telescopic tube 84, and the sealing element 85 is capable of sealing the central outlet 843 of the telescopic tube 84 and preventing the formation of a gap between the nozzle support 82 and the telescopic tube 84. The reset seat 86 is movably disposed in the deformation space 821 of the nozzle support 82, and the reset seat 86 abuts against the sealing element 85. The first reset element 87 is deformably disposed in the deformation space 821 of the nozzle support 82, and the opposite ends of the first reset element 87 abut against the inner wall of the nozzle support 82 and the reset seat 86, respectively.

[0116] When the nozzle 80 is in a non-operating state, the first reset element 87 and the reset seat 86 allow the sealing element 85 to be held in a state for closing the central outlet 843 of the telescopic tube 84.

[0117] When the nozzle 80 is needed to clean the lens surface of the vehicle-mounted camera 100, firstly, the telescopic tube 84 is driven to move relative to the sleeve 81 to allow the nozzle 83 to be exposed, with the nozzle 83 facing the lens surface of the vehicle-mounted camera 100. Secondly, the pressure of the cleaning fluid in the tube channel 841 of the telescopic tube 84 is increased to allow the cleaning fluid to apply pressure to the sealing element 85, causing the sealing element 85 to deform. At this time, on the one hand, the sealing element 85 pushes the first reset element 87 to deform through the reset seat 86, allowing the first reset element 73 to accumulate elastic potential energy. On the other hand, a liquid flow channel 801 of the nozzle 80 is formed between the sealing element 85 and the telescopic tube 84 to connect the central liquid outlet 843 of the telescopic tube 84 and the annular liquid outlet groove 844, and the... The sealing element 85 prevents the formation of a gap between the nozzle bracket 82 and the telescopic tube 84, so that the cleaning fluid in the tube channel 841 of the telescopic tube 84 can be sprayed out of the nozzle 80 and onto the lens surface of the vehicle camera 100 through the central outlet 843, the annular outlet groove 844 and the peripheral outlet through hole 845 of the telescopic tube 84, the outlet channel 823 of the nozzle bracket 82 and the nozzle body 83, thereby cleaning the contaminants adhering to the lens surface of the vehicle camera 100 and ensuring that the vehicle camera 100 can capture clear images.

[0118] When the pressure of the cleaning fluid in the tube channel 841 of the telescopic tube 84 is reduced to less than the elastic potential energy of the first reset element 87, the first reset element 87, during the process of restoring its initial state, pushes the sealing element 85 towards the telescopic tube 84 via the reset seat 86, allowing the sealing element 85 to restore the closure of the central outlet 843 of the telescopic tube 84. Furthermore, the telescopic tube 84 can be driven to move relative to the sleeve 81.

[0119] During the process of cleaning the lens surface of the vehicle-mounted camera 100 using the nozzle 80, on the one hand, the gas in the deformation space 821 of the nozzle bracket 82 can be discharged from the vent 822 of the nozzle bracket 82 as the reset seat 86 is displaced. On the other hand, the cleaning fluid in the tube channel 841 of the telescopic tube 84 will not form backflow when it is sprayed out sequentially through the central outlet 843, the annular outlet groove 844 and the peripheral outlet through hole 845 of the telescopic tube 84, the outlet channel 823 of the nozzle bracket 82 and the nozzle body 83. In this way, the pressure value of the cleaning fluid when it is sprayed out from the nozzle body 83 is the same as the pressure value of the cleaning fluid in the tube channel 841 of the telescopic tube 84, so as to ensure the spraying speed of the cleaning fluid from the nozzle 80 to the lens surface of the vehicle-mounted camera 100, thereby improving the cleaning effect of the lens surface of the vehicle-mounted camera 100. After the lens surface of the vehicle-mounted camera 100 is cleaned, external gas can enter the deformation space 821 of the nozzle bracket 82 through the vent 822 of the nozzle bracket 82 as the reset seat 86 is displaced.

[0120] Preferably, the reset seat 86 has a body space 861, and one end of the first reset element 87 is held in the body space 861 of the reset seat 86 to prevent the first reset element 87 from disengaging from the reset seat 86, thereby ensuring the reliability and stability of the nozzle 80.

[0121] It is worth mentioning that the specific structure of the first reset element 87 is not limited in this invention. For example, the first reset element 87 can be a compression spring.

[0122] Preferably, the nozzle 80 includes two nozzle bodies 83, and correspondingly, the nozzle support 82 has two liquid outlet channels 823. Each nozzle body 83 is connected to each liquid outlet channel 823 of the nozzle support 82, thereby increasing the cleaning area of ​​the nozzle 80. More preferably, the two nozzle bodies 83 of the nozzle 80 have different orientations; for example, the upper nozzle body 83 is tilted upwards, and the lower nozzle body 83 is tilted downwards, further increasing the cleaning area of ​​the nozzle 80.

[0123] Reference Appendix Figures 9 to 12B The nozzle bracket 82 further has an installation space 824, wherein one end of the telescopic tube 84 extends into the installation space 824 of the nozzle bracket 82, and the telescopic tube 84 is installed in the installation space 824 of the nozzle bracket 82 to realize the installation of the nozzle bracket 82 and the telescopic tube 84.

[0124] Continue to refer to the appendix Figures 9 to 12B The nozzle 80 further includes a mounting base 88 and a second reset element 89, wherein one end of the sleeve 81 is mounted on the mounting base 88, the second reset element 89 is fitted onto the telescopic tube 84, and the opposite ends of the second reset element 89 abut against the telescopic tube 84 and the mounting base 88, respectively.

[0125] When the telescopic tube 84 is driven to move relative to the sleeve 81 to allow the nozzle body 83 of the nozzle 80 to be exposed, the second reset element 89 deforms and accumulates elastic potential energy; when the external force driving the telescopic tube 84 is removed, the second reset element 89 can push the telescopic tube 84 to move relative to the sleeve 81 in the process of restoring the initial state.

[0126] It is worth mentioning that the specific structure of the second reset element 89 is not limited in this invention. For example, the second reset element 89 can be a compression spring.

[0127] The mounting base 88 includes a mounting ring 881 and a first mounting sleeve 882 integrally extending to one side of the mounting ring 881. The mounting ring 881 has an annular channel 8811, and the first mounting sleeve 882 has a first mounting space 8821. The annular channel 8811 of the mounting ring 881 and the first mounting space 8821 of the first mounting sleeve 882 are connected. One end of the sleeve 81 extends to the first mounting space 8821 of the first mounting sleeve 882, and the end of the sleeve 81 and the first mounting sleeve 882 are mounted to each other. The nozzle bracket 82 extends through the annular channel 8811 of the mounting ring 881 to the first mounting space 8821 of the first mounting sleeve 882. The second reset element 89 is hidden between the mounting base 88 and the sleeve 81, and the opposite ends of the second reset element 89 abut against one end of the telescopic tube 84 and the mounting ring 881, respectively.

[0128] The mounting base 88 further includes a second mounting tube 883 having a second mounting space 8831, wherein the second mounting tube 883 integrally extends to the other side of the mounting ring 881, and the second mounting space 8831 of the second mounting tube 883 communicates with the ring channel 8811 of the mounting ring 881, wherein a portion of the mouthpiece 83 can be mounted in the second mounting space 8831 of the second mounting tube 883 so that the orientation of the mouthpiece 83 is restricted by the second mounting tube 883 of the mounting base 88.

[0129] Continue to refer to the appendix Figures 9 to 12BThe nozzle 80 further includes a sealing seat 810, which is fitted onto the end of the telescopic tube 84, and the outer wall of the sealing seat 810 is fitted against the inner wall of the sleeve 81 to form a liquid-containing space 802 of the nozzle 80 between the telescopic tube 84, the sealing seat 810, and the sleeve 81. The liquid inlet 842 of the telescopic tube 84 communicates with the liquid-containing space 802. The sleeve 81 further has a liquid inlet channel 812, which is connected to the liquid-containing space 802 of the nozzle 80. When high-pressure cleaning fluid enters the liquid-containing space 802 of the nozzle 80 from the liquid inlet channel 812 of the sleeve 81, the high-pressure cleaning fluid can push the sealing seat 810 to move the telescopic tube 84 within the telescopic space 811 of the sleeve 81 relative to the sleeve 81, allowing the nozzle body 83 of the nozzle 80 to be exposed. In addition, the high-pressure cleaning fluid can cause the sealing element 85 to deform, thereby forming the liquid flow channel 801 of the nozzle 80 between the sealing element 85 and the telescopic tube 84. As a result, the cleaning fluid can be sprayed out of the nozzle 80 and onto the lens surface of the vehicle camera 100 through the central outlet 843, the annular outlet groove 844 and the peripheral outlet through hole 845 of the telescopic tube 84, the outlet channel 823 of the nozzle bracket 82 and the nozzle body 83, in order to clean the contaminants adhering to the lens surface of the vehicle camera 100, thereby allowing the vehicle camera 100 to image clearly.

[0130] The process of the nozzle 80 cleaning the lens surface of the vehicle-mounted camera 100 is as follows.

[0131] When high-pressure cleaning fluid enters the liquid-containing space 802 of the nozzle 80 from the inlet channel 812 of the sleeve 81, the high-pressure cleaning fluid pushes the sealing seat 810, causing the telescopic tube 84 to move relative to the sleeve 81 within the telescopic space 811 of the sleeve 81, allowing the nozzle body 83 of the nozzle 80 to be exposed. During this process, the second reset element 89 is deformed by being pressed by the telescopic tube 84 toward the mounting base 88 to accumulate elastic potential energy.

[0132] High-pressure cleaning fluid pushes the sealing element 85 to deform, forming the liquid flow channel 801 of the nozzle 80 between the sealing element 85 and the telescopic tube 84. This allows the cleaning fluid to be ejected sequentially from the nozzle 80 and onto the lens surface of the vehicle-mounted camera 100 through the central outlet 843, the annular outlet groove 844, the peripheral outlet through-hole 845 of the telescopic tube 84, the outlet channel 823 of the nozzle support 82, and the nozzle body 83. This cleans contaminants adhering to the lens surface of the vehicle-mounted camera 100, thereby allowing the vehicle-mounted camera 100 to capture clear images. During this process, on the one hand, the gas in the deformation space 821 of the nozzle bracket 82 can be discharged from the vent hole 822 of the nozzle bracket 82 as the reset seat 86 is displaced. On the other hand, the cleaning fluid in the tube channel 841 of the telescopic tube 84 will not form backflow when it is sprayed out sequentially through the central outlet 843, the annular outlet groove 844 and the peripheral outlet through hole 845 of the telescopic tube 84, the outlet channel 823 of the nozzle bracket 82 and the nozzle body 83. In this way, the pressure value of the cleaning fluid when it is sprayed out from the nozzle body 83 is the same as the pressure value of the cleaning fluid in the tube channel 841 of the telescopic tube 84, so as to ensure the spraying speed of the cleaning fluid from the nozzle 80 to the lens surface of the vehicle camera 100, thereby improving the cleaning effect of the lens surface of the vehicle camera 100. In other words, when the high-pressure cleaning fluid pushes the sealing element 85 to deform and flows from the central outlet 843 of the telescopic tube 84 to the annular outlet groove 844 through the liquid flow channel 801 of the nozzle 80, the pressure value of the high-pressure cleaning fluid will not be reduced, thereby ensuring the spray speed of the cleaning fluid from the nozzle 80 to the lens surface of the vehicle camera 100.

[0133] When the pressure of the cleaning fluid in the liquid-containing space 802 of the nozzle 80 decreases, under the action of the first reset element 87, the sealing element 85 returns to its initial state to prevent the central outlet 843 of the telescopic tube 84 from communicating with the annular outlet groove 844. Under the action of the second reset element 89, the telescopic tube 84 moves relative to the sleeve 81 within the telescopic space 811 of the sleeve 81.

[0134] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and explained in the embodiments, and any variations or modifications may be made to the implementation of the present invention without departing from the stated principles.

Claims

1. A method for cleaning a vehicle-mounted camera, characterized in that, The cleaning method cleans the lens surface of the vehicle camera when it is contaminated, and the cleaning method includes the following steps: (a) A telescopic tube is driven to move a nozzle bracket and a nozzle body relative to a set of tubes, so as to allow the nozzle body to be exposed and facing the lens surface of a vehicle camera, wherein a liquid inlet and a central liquid outlet of the telescopic tube are respectively connected to a tube body channel of the telescopic tube, at least one peripheral liquid outlet through hole of the telescopic tube is connected to an annular liquid outlet groove surrounding the central liquid outlet, and at least one liquid outlet channel of the nozzle bracket is respectively connected to the peripheral liquid outlet through hole of the telescopic tube and the nozzle body; (b) Increase the pressure of the cleaning fluid in the tube channel of the telescopic tube to allow a sealing element that seals the central outlet of the telescopic tube to deform toward the nozzle body and form a liquid channel between the sealing element and the telescopic tube to connect the central outlet of the telescopic tube and the annular outlet groove. At this time, the high-pressure cleaning fluid is sprayed sequentially through the central outlet of the telescopic tube, the annular outlet groove and the peripheral outlet through hole, the outlet channel of the nozzle bracket and the nozzle body onto the lens surface of the vehicle camera to clean the lens surface of the vehicle camera. as well as (c) Reduce the pressure of the cleaning fluid in the tube channel of the telescopic tube to allow the sealing element to return to its initial state in the direction of the telescopic tube and reseal the central outlet of the telescopic tube; In step (b), the sealing element is allowed to push a first reset element toward the nozzle body to deform and accumulate elastic potential energy. In step (c), the first reset element is allowed to push the sealing element toward the telescopic tube during the process of restoring the initial state, so that the sealing element is kept in the state of sealing the central liquid outlet of the telescopic tube. In step (b), the sealing element is allowed to deform by pushing the first reset element toward the nozzle body through a reset seat; in step (c), the first reset element is allowed to push the sealing element toward the telescopic tube through the reset seat. Before step (a), it is determined whether the lens surface of the vehicle-mounted camera is contaminated. If the lens surface of the vehicle-mounted camera is contaminated, steps (a) to (c) are performed to clean the lens surface of the vehicle-mounted camera. The contamination determination method for determining whether the lens surface of the vehicle-mounted camera is contaminated includes the following steps: (a) In an indoor scene platform environment, the vehicle-mounted camera is allowed to capture a poster containing scene information through contaminated specimens of different grayscale levels to obtain a series of specimen images, wherein step (a) further includes the step of: (a.1) Construct a specimen image acquisition platform indoors; (a.2) The poster is arranged at one end of the specimen image acquisition stand, and the vehicle-mounted camera is arranged at the other end of the specimen image acquisition stand; (a.3) The contaminated specimen is placed between the vehicle-mounted camera and the poster to allow the vehicle-mounted camera to photograph the poster through the contaminated specimen, wherein the contaminated specimen is replaceable; (b) Convert the raw image data of the series of specimen images into image data in HSV format; (c) The mean and variance of the HSV color space of the HSV format image data of a series of specimen images are used as features and input into the support vector machine algorithm model to train a classification algorithm model. (d) Input the image data of a detection image captured by the vehicle-mounted camera as a feature into the classification algorithm model, so that the classification algorithm model can classify the detection image captured by the vehicle-mounted camera; (e) Based on the classification result of the classification algorithm model, determine whether the lens surface of the vehicle camera is contaminated; The poster is a virtual poster projected onto the end of the specimen image acquisition platform to improve the generalization ability and robustness of the specimen image and increase the acquisition speed of the specimen image by replacing the contaminated specimen with one of different gray levels and projecting the poster with different scene information, thereby improving the efficiency of training the classification algorithm model.

2. The cleaning method for a vehicle-mounted camera according to claim 1, wherein the edge of the sealing element is held between the nozzle bracket and the telescopic tube, and in step (b), the middle portion of the sealing element is allowed to extend into the interior of a deformation space of the nozzle bracket after deformation to form the liquid channel between the sealing element and the telescopic tube.

3. The cleaning method for a vehicle-mounted camera according to claim 2, wherein the first reset element is deformably disposed in the deformation space of the nozzle bracket, wherein in step (b), when the sealing element deforms in the direction toward the nozzle body, the gas in the deformation space of the nozzle bracket is discharged from a vent hole of the nozzle bracket, and in step (c), when the sealing element deforms in the direction toward the telescopic tube, external gas enters the deformation space of the nozzle bracket from the vent hole of the nozzle bracket.

4. The cleaning method for a vehicle-mounted camera according to any one of claims 1 to 3, wherein in step (a), the two nozzles are allowed to face the lens surface of the vehicle-mounted camera in different directions.

5. The cleaning method for a vehicle-mounted camera according to any one of claims 1 to 3, wherein opposite ends of a second reset element fitted onto the telescopic tube abut against the telescopic tube and a mounting base for mounting the sleeve, wherein in step (a), when the telescopic tube is driven to move the nozzle bracket and the nozzle body relative to the sleeve, the telescopic tube and the mounting base squeeze the second reset element to deform it and accumulate elastic potential energy; after step (c), the cleaning method further includes the step: (d) allowing the second reset element to push the telescopic tube to move relative to the sleeve during the process of restoring its initial state.

6. The cleaning method for a vehicle-mounted camera according to any one of claims 1 to 3, wherein in step (a), high-pressure cleaning fluid is allowed to drive the telescopic tube to move the nozzle bracket and the nozzle body relative to the sleeve.

7. The cleaning method for a vehicle-mounted camera according to claim 5, wherein in step (a), high-pressure cleaning fluid is allowed to drive the telescopic tube to move the nozzle bracket and the nozzle body relative to the sleeve.

8. The cleaning method for a vehicle-mounted camera according to claim 7, wherein a liquid-containing space is formed between a sealing seat fitted to the end of the telescopic tube, the telescopic tube, and the sleeve, and the liquid-containing space is connected to a liquid inlet channel of the sleeve, wherein in step (a), when high-pressure cleaning fluid enters the liquid-containing space from the liquid inlet channel of the sleeve, the high-pressure cleaning fluid drives the telescopic tube to move relative to the sleeve within a telescopic space of the sleeve through the sealing seat.

Images