Vehicle camera cleaning control method, system, vehicle and medium

Abstract

The application relates to the technical field of intelligent auxiliary driving, and discloses a vehicle-mounted camera cleaning control method, a vehicle-mounted camera cleaning control system, a vehicle and a medium. The vehicle-mounted camera cleaning control method comprises the following steps: obtaining a target driving function instruction triggered by a user and initial feature points related to the target driving function instruction in a real-time image; performing screening on the initial feature points according to a function feature library corresponding to the target driving function instruction to obtain actual recognition feature points; determining a feature point recognition rate of the vehicle executing the target driving function instruction based on the actual recognition feature points and the function feature library; and determining whether to clean the vehicle-mounted camera according to the feature point recognition rate and image parameters of the real-time image. The vehicle-mounted camera cleaning control method can automatically link the vehicle-mounted camera cleaning in the intelligent auxiliary driving scene, does not need manual intervention of the user, and improves the user experience.

Description

Technical Field

[0001] This application relates to the field of intelligent assisted driving technology, and in particular to a method, system, vehicle, and medium for controlling the cleaning of an in-vehicle camera. Background Technology

[0002] In the development of intelligent assisted driving technology, the cleanliness of the camera lens, as a core environmental perception component, directly determines the success rate of the intelligent assisted driving function's activation and operational safety. In practice, camera contamination caused by rain, dust, oil, etc., is a frequent cause of malfunctions that prevent the intelligent assisted driving function from triggering properly. Existing related technologies mainly focus on two categories: one is the improvement of camera cleaning hardware structure, such as ultrasonic cleaning and high-pressure water spray structures; the other is non-scenario-based camera contamination self-checks, which periodically or randomly detect and clean contamination. Existing related technologies mainly have the following problems: when the user triggers the intelligent assisted driving function, the camera cannot be automatically cleaned, resulting in cumbersome operation and a poor user experience. Summary of the Invention

[0003] In view of this, embodiments of this application provide a vehicle-mounted camera cleaning control method, system, vehicle, and medium, which can effectively improve the problems of cumbersome operation and poor user experience caused by the inability to automatically link camera cleaning when the user triggers the intelligent assisted driving function.

[0004] In a first aspect, embodiments of this application provide a method for controlling the cleaning of an in-vehicle camera, including: Acquire initial feature points related to the target driving function command triggered by the user and in real-time images captured by the vehicle camera; The initial feature points are filtered according to the functional feature library corresponding to the target driving function command to obtain the actual identification feature points necessary for the vehicle to execute the target driving function command; Based on the actual identified feature points and the functional feature library, the feature point recognition rate of the vehicle executing the target driving function command is determined; Whether to clean the vehicle-mounted camera is determined based on the feature point recognition rate and the image parameters of the real-time image.

[0005] In a first possible embodiment of the first aspect, before filtering the initial feature points according to the functional feature library corresponding to the target driving function instruction, the method further includes: Determine the functional feature points required for the vehicle to execute the target driving function command, and construct a functional feature library corresponding to the target driving function command based on the functional feature points. The functional feature library includes all functional feature points necessary for the vehicle to execute the target driving function command. The step of filtering the initial feature points according to the functional feature library corresponding to the target driving function command includes: The initial feature points are matched with each functional feature point in the functional feature library; Filter out initial feature points that do not match the functional feature library, and retain the initial feature points that match the functional feature library to obtain the actual recognition feature points.

[0006] In a second possible embodiment of the first aspect, determining the feature point recognition rate of the vehicle executing the target driving function command based on the actual identified feature points and the functional feature library includes: Determine the number of first feature points of the actual identified feature points shown and the number of second feature points of all functional feature points in the functional feature library; The feature point recognition rate is obtained by calculating the ratio of the number of the first feature points to the number of the second feature points.

[0007] In a third possible embodiment of the first aspect, determining whether to clean the vehicle-mounted camera based on the feature point recognition rate and the image parameters of the real-time image includes: Based on the comparison result between the feature point recognition rate and the preset recognition rate threshold, it is determined whether the vehicle executes the target driving function command abnormally; Based on the comparison results between the image parameters and the corresponding preset image parameters, it is determined whether the vehicle-mounted camera is dirty; If the target driving function command is executed abnormally and the vehicle camera is dirty, the vehicle camera shall be cleaned.

[0008] In a fourth possible embodiment of the first aspect, determining whether the vehicle's execution of the target driving function command is abnormal based on the comparison result of the feature point recognition rate and a preset recognition rate threshold includes: If the feature point recognition rate is greater than or equal to the preset recognition rate threshold, it is determined that the target driving function command is being executed normally. If the feature point recognition rate is less than the preset recognition rate threshold, the execution of the target driving function command is determined to be abnormal.

[0009] In a fifth possible embodiment of the first aspect, the image parameters include image contrast, image sharpness, and the proportion of image occlusion area. Determining whether the vehicle-mounted camera is dirty based on a comparison of the image parameters with corresponding preset image parameters includes: If the image contrast is greater than or equal to a preset contrast, the image sharpness is greater than or equal to a preset sharpness, and the proportion of the image occluded area is less than or equal to a preset occluded area, then the vehicle camera is determined to be free of dirt. If the image contrast is less than the preset contrast, the image sharpness is less than the preset sharpness, or the proportion of the image occluded area is greater than the preset occluded area proportion, the vehicle camera is determined to be dirty.

[0010] In a sixth possible embodiment of the first aspect, it further includes: If the target driving function command is executed abnormally and the vehicle camera is not dirty, output information indicating that the vehicle is executing the target driving function command abnormally. If the target driving function command is executed normally and the vehicle camera is not dirty, then the target driving function command is executed. If the target driving function command is executed normally and the vehicle camera is dirty, the vehicle camera will not be cleaned before the target driving function command is executed.

[0011] Secondly, embodiments of this application provide a vehicle-mounted camera cleaning control system, comprising: The data acquisition module is used to acquire the target driving function command triggered by the user and the initial feature points related to the target driving function command in the real-time image captured by the vehicle camera; The feature point filtering module is used to filter the initial feature points according to the functional feature library corresponding to the target driving function command, so as to obtain the actual identification feature points necessary for the vehicle to execute the target driving function command. The recognition rate calculation module is used to determine the feature point recognition rate of the vehicle executing the target driving function command based on the actual recognized feature points and the functional feature library; The cleaning control module is used to determine whether to clean the vehicle-mounted camera based on the feature point recognition rate and the image parameters of the real-time image.

[0012] Thirdly, embodiments of this application provide a vehicle, the vehicle including a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the above-described vehicle-mounted camera cleaning control method.

[0013] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described vehicle-mounted camera cleaning control method.

[0014] The embodiments of this application have the following beneficial effects: This embodiment of a vehicle-mounted camera cleaning control method includes: acquiring a user-triggered target driving function command and initial feature points related to the target driving function command in a real-time image captured by the vehicle-mounted camera; filtering the initial feature points according to a functional feature library corresponding to the target driving function command to obtain the actual recognition feature points necessary for the vehicle to execute the target driving function command; determining the feature point recognition rate for the vehicle to execute the target driving function command based on the actual recognition feature points and the functional feature library; and determining whether to clean the vehicle-mounted camera based on the feature point recognition rate and image parameters of the real-time image. Based on the above scheme, this vehicle-mounted camera cleaning control method can detect whether to link vehicle-mounted camera cleaning in intelligent assisted driving scenarios based on the feature point recognition rate and image parameters of the real-time image, thereby achieving automatic linkage for vehicle-mounted camera cleaning without manual user intervention, improving user experience. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This paper shows a schematic diagram of a first embodiment of the vehicle-mounted camera cleaning control method of this application; Figure 2 This paper illustrates a second flowchart of the vehicle-mounted camera cleaning control method according to an embodiment of this application. Figure 3 This paper illustrates a third flowchart of the vehicle-mounted camera cleaning control method according to an embodiment of this application. Figure 4 A schematic diagram of a vehicle-mounted camera cleaning control system according to an embodiment of this application is shown.

[0017] Explanation of key component symbols: 200 - Vehicle-mounted camera cleaning control system; 210 - Data acquisition module; 220 - Feature point filtering module; 230 - Recognition rate calculation module; 240 - Cleaning control module. Detailed Implementation

[0018] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0019] The components of the embodiments of this application described and illustrated in the accompanying drawings can be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] In the following text, the terms "comprising," "having," and their cognates, which may be used in various embodiments of this application, are intended only to indicate a particular feature, number, step, operation, element, component, or combination thereof, and should not be construed as primarily excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, or adding the possibility of one or more combinations thereof. Furthermore, the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0021] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of this application pertain. Terms (such as those defined in commonly used dictionaries) shall be interpreted as having the same meaning as in their contextual meaning in the relevant technical field and shall not be construed as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of this application.

[0022] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0023] The following describes the vehicle-mounted camera cleaning control method using specific embodiments.

[0024] Figure 1 A flowchart of a vehicle-mounted camera cleaning control method according to an embodiment of this application is shown. Exemplarily, the vehicle-mounted camera cleaning control method includes the following steps: S110: Obtain the initial feature points related to the target driving function command triggered by the user and the real-time image captured by the vehicle camera.

[0025] In this embodiment, the target driving function command is an intelligent assisted driving function trigger command sent by the user through the in-vehicle terminal (such as the central control touch screen, voice command, physical button, etc.). This intelligent assisted driving function trigger command includes, but is not limited to, parking assist function commands, remote parking function commands, memory parking function commands, and assisted driving function commands. The real-time image is one or more frames captured by the in-vehicle camera during the execution of the target driving function command after the user triggers it.

[0026] As an example, after a target driving function command is triggered, the corresponding function needs to be executed based on real-time images captured by the vehicle's camera. For instance, after a parking assist function command is triggered, initial feature points required for parking assist need to be extracted based on real-time images captured by the vehicle's camera, such as parking space corner points, parking space border lines, parking locks, limiters, and parking arrows on the road surface. The initial feature points are extracted from the real-time image using corner detection algorithms and edge detection algorithms, focusing on feature points related to the execution of the target driving function command.

[0027] In one embodiment, initial feature point extraction relies on edge detection algorithms to locate line contours, followed by corner detection algorithms to locate line intersections and inflection points. In intelligent assisted driving scenarios, effective feature points such as parking space border corners, parking line intersections, and border lines are specifically extracted to provide foundational data for intelligent assisted driving functions such as parking space positioning, coordinate calculation, and parking path planning. In this embodiment, the original real-time image contains noise, uneven brightness, and glare interference. Before feature point extraction, the real-time image is preprocessed to improve the stability of feature point extraction. Preprocessing includes, but is not limited to, filtering and noise reduction, and contrast enhancement.

[0028] S120: Based on the functional feature library corresponding to the target driving function command, the initial feature points are filtered to obtain the actual identification feature points necessary for the vehicle to execute the target driving function command.

[0029] For example, the initial feature points include feature points related to the target driving function command. The initial feature points need to be filtered to obtain the actual recognition feature points necessary for the target driving function command extracted based on the real-time image.

[0030] In one embodiment, before filtering the initial feature points according to the functional feature library corresponding to the target driving function command, the functional feature points required for the vehicle to execute the target driving function command are determined, and the functional feature library corresponding to the target driving function command is constructed based on the functional feature points. The functional feature library includes all functional feature points necessary for the vehicle to execute the target driving function command.

[0031] In this embodiment, when a vehicle executes a target driving function command, it first needs to rely on the camera position (such as parking assistance relying on the rear camera), and secondly on the functional feature points required by the target driving function command. Different driving function commands have significantly different requirements for functional feature points, which is the core basis for subsequent cleaning determination of the vehicle camera.

[0032] For example, in one embodiment, the functional feature points required for parking assistance function commands include, but are not limited to, parking space corner points, parking space border lines, parking locks, limiters, and road parking arrows, among which parking space corner points and parking space border lines are core mandatory feature points, and the rest are auxiliary feature points. The functional feature points required for remote parking function commands include, but are not limited to, obstacle outlines, parking space edges, the drivable area around the vehicle, and curbs, with a focus on the recognition accuracy of near-distance feature points. The functional feature points required for memory parking function commands include, but are not limited to, path feature points (such as intersection turning points and lane change markers), static environmental markers (such as fixed streetlights and road signs), and lane lines, ensuring the continuity and stability of feature points. The functional feature points required for assisted driving function commands include, but are not limited to, lane lines, pedestrians, vehicles, traffic lights, speed limit signs, and zebra crossings, balancing the recognition effect of dynamic feature points (pedestrians and vehicles) and static feature points.

[0033] In this embodiment, for each driving function command, the functional feature points necessary for the normal activation of the driving function command are preset, and a functional feature library corresponding to each driving function command is constructed as the benchmark for subsequent calculation of feature point recognition rate. For example, if the parking assistance function command recognizes at least four parking space line corner points, the front edge line of the parking space, and the rear edge line of the parking space, then the corresponding functional feature library includes the four parking space line corner points, the front edge line of the parking space, and the rear edge line of the parking space, and the number of second feature points for all functional feature points in the functional feature library is six. It can be understood that the functional feature libraries corresponding to the remote parking function command, the memory parking function command, and the assisted driving function command can be set according to the feature points necessary for the actual execution of the intelligent assisted driving function command, and are not limited here.

[0034] In another embodiment, the initial feature points are matched with each functional feature point in the functional feature library; the initial feature points that do not match the functional feature library are filtered out, and the initial feature points that match the functional feature library are retained to obtain the actual recognition feature points.

[0035] In this embodiment, the initial feature points extracted from the real-time image may contain redundant or irrelevant features. Using the functional feature library corresponding to the target driving function command as a matching benchmark, the initial feature points are matched one-to-one with the functional feature points in this library. Initial feature points that do not match are actively discarded, retaining only those that strictly correspond to the functional feature points in the library, thus outputting actual recognition feature points highly correlated with the functional feature library. For example, when a user triggers a parking assist function command, the extracted initial feature points may include the front and left border lines of the parking space. In this case, the front border line matches the functional feature library corresponding to the parking assist function command, while the left border line does not. Therefore, the left border line is filtered out, and only the front border line is retained. By filtering the initial feature points, fuzzy and misidentified feature points can be effectively eliminated, significantly improving the accuracy of subsequent cleaning determination, feature point recognition rate calculation, and function triggering decisions.

[0036] S130 determines the feature point recognition rate of the vehicle executing the target driving function command based on the actual identified feature points and the functional feature library.

[0037] For example, the feature point recognition rate is a key quantitative indicator for measuring whether the camera's perception capability meets the requirements for activating the target driving function. The feature point recognition rate directly reflects the impact of camera dirt on the successful execution of the target driving function command and is the core basis for determining whether the vehicle camera needs to be cleaned.

[0038] In one embodiment, the number of first feature points and the number of second feature points of all functional feature points in the functional feature library are determined; the ratio of the number of first feature points to the number of second feature points is calculated to obtain the feature point recognition rate. The formula for calculating the feature point recognition rate can be expressed as: ; In the formula, Indicates the feature point recognition rate. This indicates the number of first feature points actually identified. For example, if the actual identified feature points include two parking space corner points and the front border line of the parking space, then the number of first feature points is three. This indicates the number of second feature points for all functional feature points in the functional feature library.

[0039] For example, in one embodiment, the functional feature library corresponding to the parking assistance function command includes six core functional feature points. If three actual identification feature points are successfully extracted and matched in real time, the feature point recognition rate is 50%, which can be considered as dirt affecting the normal start-up of the function.

[0040] Optionally, during vehicle movement, the feature point recognition rate will dynamically fluctuate with changes in viewing angle, distance, and area obscured by dirt. Therefore, the feature point recognition rate of multiple frames can be calculated during vehicle movement, and the average of the recognition rates of each feature point can be taken to obtain the average feature point recognition rate. The average feature point recognition rate is used as the basis for subsequent judgment to ensure stable and reliable judgment.

[0041] S140 determines whether to clean the vehicle-mounted camera based on the feature point recognition rate and the image parameters of the real-time image.

[0042] For example, the image parameters of a real-time image are quantitative indicators that characterize the imaging quality of a camera, including but not limited to image contrast, image sharpness, and the percentage of occluded areas in the image.

[0043] When the real-time image includes a single frame captured by the vehicle camera during the execution of the target driving function command, only the image parameters of that frame are selected. When the real-time image includes multiple frames captured by the vehicle camera during the execution of the target driving function command, the average image parameters of the multiple frames are selected. That is, the real-time image in this case includes, but is not limited to, average image contrast, average image sharpness, and average occlusion area percentage. The average image contrast is the average contrast of all frames, the average image sharpness is the average sharpness of all frames, and the average occlusion area percentage is the average occlusion area of ​​all frames. By combining image parameters and feature point recognition rate, a two-layer joint judgment logic is constructed to avoid the shortcomings of single-index judgment.

[0044] In one embodiment, such as Figure 2 As shown, determining whether to clean the vehicle camera involves the following steps: S141, Based on the comparison result between the feature point recognition rate and the preset recognition rate threshold, determine whether the vehicle is abnormally executing the target driving function command.

[0045] In this embodiment, the preset recognition rate threshold is a minimum feature point recognition rate benchmark value set for different driving function commands to ensure their safe and reliable activation. Different preset recognition rate thresholds are set according to different driving function commands. For example, the preset recognition rate threshold for parking assist function commands is 80%, the preset recognition rate threshold for remote parking function commands is 75%, and the preset recognition rate threshold for memory parking function commands is 85%.

[0046] In one embodiment, if the feature point recognition rate is greater than or equal to a preset recognition rate threshold, the target driving function command is determined to be executed normally; if the feature point recognition rate is less than the preset recognition rate threshold, the target driving function command is determined to be executed abnormally.

[0047] In this embodiment, when the target driving function command is executed normally, the vehicle can execute the target driving function command based on the feature points related to the target driving function in the real-time image; when the target driving function command is executed abnormally, the vehicle cannot execute the target driving function command based on the feature points related to the target driving function in the real-time image, and it is necessary to further determine whether the vehicle camera is dirty.

[0048] S142, Based on the comparison result between the image parameters and the corresponding preset image parameters, determine whether the vehicle camera is dirty.

[0049] For example, the preset image parameters are the normal imaging lower or upper limits set for each image parameter. They can be obtained statistically from uncontaminated calibration images and used as a benchmark threshold for judging lens contamination, fogging, or foreign object obstruction.

[0050] In one embodiment, the vehicle camera is determined to be clean when the image contrast is greater than or equal to a preset contrast, the image sharpness is greater than or equal to a preset sharpness, and the proportion of the image occluded area is less than or equal to a preset occluded area proportion; the vehicle camera is determined to be dirty when one or more of the following conditions exist: the image contrast is less than a preset contrast, the image sharpness is less than a preset sharpness, and the proportion of the image occluded area is greater than a preset occluded area proportion.

[0051] In this embodiment, a multi-dimensional image parameter joint determination of the camera's dirt status is employed. Any parameter exceeding its limit triggers a dirt determination, improving the accuracy of dirt detection for vehicle-mounted cameras. For example, a preset contrast ratio of 50, a preset sharpness of 60, and a preset occlusion area ratio of 30% are set. If all these conditions are met—image contrast greater than or equal to 50, image sharpness greater than or equal to 60, and occlusion area ratio less than or equal to 30%—the image parameters meet the standards, and the vehicle-mounted camera is not dirty; otherwise, the image parameters do not meet the standards, and the vehicle-mounted camera is dirty.

[0052] S143, If the target driving function command is executed abnormally and the vehicle camera is dirty, clean the vehicle camera.

[0053] In this embodiment, the cleaning action is only initiated when both the feature point recognition function malfunction (the target driving function command cannot be executed normally) and the camera is dirty (imaging layer degradation) are simultaneously detected, thus avoiding ineffective cleaning caused by misjudgment of a single indicator. When both the feature point recognition function malfunction and the camera are dirty, the detection of camera dirt causing the intelligent assisted driving function to fail to start automatically cleans the vehicle camera, enabling the intelligent assisted driving function to start successfully and improving the success rate of intelligent assisted driving startup.

[0054] In one implementation, if it is determined that a dirty camera is causing abnormal execution of the target driving function commands, precise fault alerts and handling instructions can be displayed on the in-vehicle terminal. This can be done using a multimodal approach combining text and voice, and the alert content incorporates the feature point recognition rate and the cleaning progress of the in-vehicle camera, allowing the user to clearly understand the details of the abnormality and improving the user experience. For example, the alert message could be: "Camera is dirty, feature point recognition rate is only 50%, parking assist cannot be activated, cleaning camera in progress [cleaning progress: 0%]".

[0055] In one embodiment, when cleaning the vehicle-mounted camera, the cleaning parameters can be controlled based on the proportion of the image occlusion area, image contrast, image sharpness, and feature point recognition rate. These parameters include, but are not limited to, the number of water spray cycles, the number of swiping cycles, the cleaning duration, and the water spray pressure, to ensure effective cleaning. A lower proportion of the image occlusion area corresponds to fewer water spray cycles, fewer swiping cycles, a shorter cleaning duration, and lower water spray pressure. Lower image contrast corresponds to more water spray cycles, more swiping cycles, a longer cleaning duration, and higher water spray pressure. Lower image sharpness corresponds to more water spray cycles, more swiping cycles, a longer cleaning duration, and higher water spray pressure. A lower feature point recognition rate corresponds to fewer water spray cycles, fewer swiping cycles, a shorter cleaning duration, and lower water spray pressure.

[0056] For example, when the occlusion rate is less than or equal to 10%, the feature point recognition rate is 60%-80%. At this time, dirt has little impact on the feature point recognition rate. A "single water spray + single brushing" cleanup is performed, lasting 1-2 seconds, with moderate water pressure to avoid splashing water and affecting other cameras. When the occlusion rate is 10%-30%, the feature point recognition rate is 40%-60%, and dirt significantly affects feature point recognition. A "two-cycle water spray + brushing" cleanup is performed, lasting 3-4 seconds, with slightly higher water pressure to ensure removal of attached dust and light oil. When the occlusion rate is greater than 30%, the feature point recognition rate is less than 40%, indicating severe dirt and most feature points cannot be recognized. A "three-cycle water spray + brushing + air drying" cleanup is performed, lasting 5-6 seconds, with maximum water pressure. The air drying step prevents residual water stains on the lens from affecting feature point recognition after cleaning.

[0057] In another embodiment, after cleaning is completed, the image clarity is not directly judged. Instead, the complete vehicle camera cleaning judgment process of S110~S140 is re-executed to ensure that the cleaning effect can meet the function start-up requirements.

[0058] The specific process is as follows: In the re-inspection step, the actual recognition feature points necessary for the target driving function command are re-extracted, the feature point recognition rate is recalculated, and image parameters are simultaneously detected, performing a two-layer joint judgment. If the re-inspection is successful, a prompt will be displayed on the vehicle terminal stating "Cleaning complete, feature point recognition rate meets standard, function is starting." The vehicle domain controller automatically triggers the target driving function command, eliminating the need for secondary user intervention and achieving automated "cleaning-re-inspection-retry." If the re-inspection fails, a secondary reminder will be displayed on the vehicle terminal stating "Dirty residue remains after camera cleaning, feature point recognition rate does not meet standard, function cannot be started, please clean manually." Simultaneously, the current function triggering process will be terminated, and the degree of dirt, cleaning parameters, and re-inspection results will be recorded to provide data support for subsequent cleaning parameter optimization.

[0059] In one embodiment, such as Figure 3 As shown, the vehicle-mounted camera cleaning control method also includes the following steps: S150 outputs information indicating an abnormality in the execution of the target driving function command, provided that the vehicle camera is not dirty.

[0060] In this embodiment, when the target driving function command is executed abnormally and the vehicle camera is not dirty, the abnormality is not caused by dirt on the vehicle camera. Therefore, the vehicle camera does not need to be cleaned to avoid ineffective cleaning. Instead, the vehicle terminal can display the abnormality information of the target driving function command, such as outputting a prompt "Function cannot be started, please check the environment" to prompt the user to check in time.

[0061] S160 executes the target driving function command if the target driving function command is executed normally and the vehicle camera is not dirty.

[0062] In this embodiment, when the target driving function command is executed normally and the vehicle camera is not dirty, the risk of abnormal execution of the target driving function command and the risk of the vehicle camera being dirty are eliminated, and the target driving function command can be executed normally.

[0063] S170: If the target driving function command is executed normally and the vehicle camera is dirty, the target driving function command will not be cleaned.

[0064] In this embodiment, when the target driving function command is executed normally, but the vehicle camera is dirty, the vehicle camera does not need to be cleaned, because at this time the dirt on the vehicle camera does not affect the normal execution of the target driving function, thus avoiding ineffective cleaning of the vehicle camera.

[0065] For example, in one embodiment, the vehicle includes a vehicle body, an intelligent assisted driving camera group (front-view, side-view, rear-view, etc.), a cleaning device corresponding to each camera, and an in-vehicle terminal (central control screen, voice module), wherein: the camera group is set in key areas for intelligent assisted driving perception such as the windshield, rear bumper, and left and right rearview mirrors; the cleaning device is precisely aligned with the cameras and has functions of spraying water, wiping, and drying; the in-vehicle terminal is linked with a multimodal interactive reminder module to realize user interaction.

[0066] With existing technology, when a vehicle approaches a parking space and the user triggers the parking assist function, if the rear-view fisheye camera has localized dirt, the current technology only detects global image parameters and mistakenly judges the camera as normal. After the parking assist function is activated, it fails to recognize enough parking space line feature points, causing the function to fail midway and only outputting a vague "parking assist failed" message. The user cannot pinpoint the cause and needs to manually check the camera, clean it, and re-trigger the function.

[0067] In this application, the user triggers a parking assist function command. This function command relies on a camera, and the function feature library includes six function feature points, with a corresponding preset recognition rate threshold of 80%. Real-time images from the camera are acquired, and two parking space line corner points are extracted. At this point, the feature point recognition rate is 33%, and the occlusion rate in the image parameters is 15%, indicating the vehicle camera is dirty. Through a dual-layer joint judgment, it is determined that the dirty camera is causing the parking assist function command to fail to execute properly. At this time, a prompt message can be immediately output: "Camera is dirty, function adaptation feature point recognition rate is only 33%, parking assist cannot be started, cleaning your camera [Cleaning progress: 0%]", simultaneously triggering the vehicle camera cleaning process. At this point, the occlusion rate is 15%, the recognition rate is 33%, and "two cycles of water spray + scraping" are performed, with a cleaning time of 3 seconds. After cleaning, during a re-inspection, six parking space line corner points are extracted, the function adaptation feature point recognition rate increases to 100%, the image parameters meet the standards, and the parking assist function can be automatically activated without secondary user operation, achieving rapid function recovery and significantly improving user experience and functional reliability.

[0068] In this embodiment, the misconception that image quality is equivalent to functional availability in the prior art is completely broken. By taking the feature point recognition rate as the core, a strong binding between dirt determination and function execution results is achieved, which not only avoids ineffective cleaning, but also eliminates the safety hazard of images meeting the standards but functions being unusable.

[0069] Figure 4 A schematic diagram of a vehicle-mounted camera cleaning control system 200 according to an embodiment of this application is shown. Exemplarily, the vehicle-mounted camera cleaning control system 200 includes: The data acquisition module 210 is used to acquire the target driving function command triggered by the user and the initial feature points related to the target driving function command in the real-time image captured by the vehicle camera.

[0070] The feature point filtering module 220 is used to filter the initial feature points according to the functional feature library corresponding to the target driving function command, so as to obtain the actual recognition feature points necessary for the vehicle to execute the target driving function command.

[0071] In one embodiment, the feature point filtering module 220 is further configured to determine the functional feature points required for the vehicle to execute the target driving function command, and construct a functional feature library corresponding to the target driving function command based on the functional feature points. The functional feature library includes all functional feature points necessary for the vehicle to execute the target driving function command. Initial feature points are matched with each functional feature point in the functional feature library; initial feature points that do not match the functional feature library are filtered out, and initial feature points that match the functional feature library are retained to obtain the actual identified feature points.

[0072] The recognition rate calculation module 230 is used to determine the feature point recognition rate of the vehicle executing the target driving function command based on the actual recognized feature points and the functional feature library.

[0073] In one embodiment, the recognition rate calculation module 230 is further configured to determine the number of first feature points of the actual recognized feature points and the number of second feature points of all functional feature points in the functional feature library; calculate the ratio of the number of first feature points to the number of second feature points to obtain the feature point recognition rate.

[0074] The cleaning control module 240 is used to determine whether to clean the vehicle-mounted camera based on the feature point recognition rate and the image parameters of the real-time image.

[0075] In one embodiment, the cleaning control module 240 is further configured to determine whether the vehicle is abnormally executing the target driving function command based on the comparison result of the feature point recognition rate and the preset recognition rate threshold; determine whether the vehicle camera is dirty based on the comparison result of the image parameters and the corresponding preset image parameters; and clean the vehicle camera if the target driving function command is abnormally executed and the vehicle camera is dirty.

[0076] In one embodiment, the cleaning control module 240 is further configured to determine that the target driving function command is executed normally when the feature point recognition rate is greater than or equal to a preset recognition rate threshold; and to determine that the target driving function command is executed abnormally when the feature point recognition rate is less than the preset recognition rate threshold.

[0077] In one embodiment, the image parameters include image contrast, image sharpness, and the percentage of the image occluded area. The cleaning control module 240 is further configured to determine that the vehicle camera is not dirty when the image contrast is greater than or equal to a preset contrast, the image sharpness is greater than or equal to a preset sharpness, and the percentage of the image occluded area is less than or equal to a preset occluded area percentage; and to determine that the vehicle camera is dirty when there is one or more of the following: the image contrast is less than a preset contrast, the image sharpness is less than a preset sharpness, and the percentage of the image occluded area is greater than a preset occluded area percentage.

[0078] In one embodiment, the cleaning control module 240 is further configured to output abnormal information about the vehicle's execution of the target driving function command when the target driving function command is executed abnormally and the vehicle camera is not dirty; execute the target driving function command when the target driving function command is executed normally and the vehicle camera is not dirty; and execute the target driving function command without cleaning the vehicle camera when the target driving function command is executed normally and the vehicle camera is dirty.

[0079] It is understood that the system in this embodiment corresponds to the vehicle camera dirt detection method in the above embodiment, and the options in the above embodiment are also applicable to this embodiment, so they will not be described again here.

[0080] This application also provides a vehicle, exemplary in that the vehicle includes a processor and a memory, wherein the memory stores a computer program, and the processor, by running the computer program, causes the vehicle to perform the functions of the various modules in the above-described vehicle-mounted camera dirt detection method or the above-described vehicle-mounted camera cleaning control system.

[0081] The processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including at least one of a Central Processing Unit (CPU), Graphics Processing Unit (GPU), Network Processor (NP), Digital Signal Processor (DSP), Application-Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0082] Memory can be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), and Electrically Erasable Programmable Read-Only Memory (EEPROM). Memory is used to store computer programs, and the processor can execute these programs upon receiving execution instructions.

[0083] This application also provides a computer-readable storage medium for storing the computer program used in the aforementioned vehicle. For example, the computer-readable storage medium may include, but is not limited to, various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0084] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that, in alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0085] In addition, the functional modules or units in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0086] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0087] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for controlling the cleaning of a vehicle-mounted camera, characterized in that, include: Acquire initial feature points related to the target driving function command triggered by the user and in real-time images captured by the vehicle camera; The initial feature points are filtered according to the functional feature library corresponding to the target driving function command to obtain the actual identification feature points necessary for the vehicle to execute the target driving function command; Based on the actual identified feature points and the functional feature library, the feature point recognition rate of the vehicle executing the target driving function command is determined; Whether to clean the vehicle-mounted camera is determined based on the feature point recognition rate and the image parameters of the real-time image.

2. The vehicle-mounted camera cleaning control method according to claim 1, characterized in that, Before filtering the initial feature points according to the functional feature library corresponding to the target driving function command, the method further includes: Determine the functional feature points required for the vehicle to execute the target driving function command, and construct a functional feature library corresponding to the target driving function command based on the functional feature points. The functional feature library includes all functional feature points necessary for the vehicle to execute the target driving function command. The step of filtering the initial feature points according to the functional feature library corresponding to the target driving function command includes: The initial feature points are matched with each functional feature point in the functional feature library; Filter out initial feature points that do not match the functional feature library, and retain the initial feature points that match the functional feature library to obtain the actual recognition feature points.

3. The vehicle-mounted camera cleaning control method according to claim 1, characterized in that, The step of determining the feature point recognition rate of the vehicle executing the target driving function command based on the actual identified feature points and the functional feature library includes: Determine the number of first feature points of the actual identified feature points shown and the number of second feature points of all functional feature points in the functional feature library; The feature point recognition rate is obtained by calculating the ratio of the number of the first feature points to the number of the second feature points.

4. The vehicle-mounted camera cleaning control method according to claim 1, characterized in that, The step of determining whether to clean the vehicle-mounted camera based on the feature point recognition rate and the image parameters of the real-time image includes: Based on the comparison result between the feature point recognition rate and the preset recognition rate threshold, it is determined whether the vehicle executes the target driving function command abnormally; Based on the comparison results between the image parameters and the corresponding preset image parameters, it is determined whether the vehicle-mounted camera is dirty; If the target driving function command is executed abnormally and the vehicle camera is dirty, the vehicle camera shall be cleaned.

5. The vehicle-mounted camera cleaning control method according to claim 4, characterized in that, The step of determining whether the vehicle's execution of the target driving function command is abnormal based on the comparison result between the feature point recognition rate and the preset recognition rate threshold includes: If the feature point recognition rate is greater than or equal to the preset recognition rate threshold, it is determined that the target driving function command is being executed normally. If the feature point recognition rate is less than the preset recognition rate threshold, the execution of the target driving function command is determined to be abnormal.

6. The vehicle-mounted camera cleaning control method according to claim 4, characterized in that, The image parameters include image contrast, image sharpness, and the percentage of occluded areas. Determining whether the vehicle-mounted camera is dirty based on a comparison of these image parameters with corresponding preset image parameters includes: If the image contrast is greater than or equal to a preset contrast, the image sharpness is greater than or equal to a preset sharpness, and the proportion of the image occluded area is less than or equal to a preset occluded area, then the vehicle camera is determined to be free of dirt. If the image contrast is less than the preset contrast, the image sharpness is less than the preset sharpness, or the proportion of the image occluded area is greater than the preset occluded area proportion, the vehicle camera is determined to be dirty.

7. The vehicle-mounted camera cleaning control method according to claim 4, characterized in that, Also includes: If the target driving function command is executed abnormally and the vehicle camera is not dirty, output information indicating that the vehicle is executing the target driving function command abnormally. If the target driving function command is executed normally and the vehicle camera is not dirty, then the target driving function command is executed. If the target driving function command is executed normally and the vehicle camera is dirty, the vehicle camera will not be cleaned before the target driving function command is executed.

8. A vehicle-mounted camera cleaning control system, characterized in that, include: The data acquisition module is used to acquire initial feature points related to the target driving function from the user-triggered target driving function command and the real-time images captured by the vehicle camera; The feature point filtering module is used to filter the initial feature points according to the functional feature library corresponding to the target driving function command, so as to obtain the actual identification feature points necessary for the vehicle to execute the target driving function command. The recognition rate calculation module is used to determine the feature point recognition rate of the vehicle executing the target driving function command based on the actual recognized feature points and the functional feature library; The cleaning control module is used to determine whether to clean the vehicle-mounted camera based on the feature point recognition rate and the image parameters of the real-time image.

9. A vehicle, characterized in that, The vehicle includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the vehicle-mounted camera cleaning control method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The device contains a computer program that, when executed by a processor, implements the vehicle camera cleaning control method as described in any one of claims 1-7.

Images