Visual judgment method, system and equipment and storage medium

Abstract

The invention belongs to the field of battery swap stations, and particularly relates to a visual judgment method, system and device and a storage medium, and the visual judgment method comprises the steps: collecting image data of a channel and an operation area in a battery swap station in real time; comparing the image data with pre-stored reference image data to obtain a target detection result; judging whether a vehicle and / or a person exists or not based on the target detection result and in combination with the environment feature information; if the vehicle and / or the person exists, the roller shutter door is controlled to be opened in a linkage mode, and a safety prompt is sent out; and if no vehicle and / or personnel exist, the roller shutter door is controlled to be closed in a linkage mode.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of battery swap stations, in particular to a visual determination method, system, device and storage medium. BACKGROUND

[0002] To ensure the efficient operation and safety of the battery swap station, the state of the passageway and work area in the station needs to be monitored in real time to determine whether there are vehicles and / or personnel, and accordingly control the related equipment (such as the roller shutter door) to ensure safety and energy saving.

[0003] The monitoring and state determination of the existing battery swap station are mainly through image acquisition and analysis and comparison. When disturbed by the environment, the imaging effect of the image is affected, resulting in deviation of the analysis and determination result, and thus the accuracy of the roller shutter door control cannot be guaranteed. SUMMARY

[0004] To solve the above problems in the prior art, the present application provides a visual determination method, system, device and storage medium.

[0005] The first aspect of the present application provides a visual determination method, comprising: real-time acquisition of image data of the passageway and work area in the battery swap station; comparing the image data with pre-stored reference image data to obtain a target detection result; based on the target detection result and combined with environmental feature information, determining whether there are vehicles and / or personnel; if there are vehicles and / or personnel, the roller shutter door is controlled to remain open and a safety prompt is issued; if there are no vehicles and / or personnel, the roller shutter door is controlled to close.

[0006] In an embodiment, before comparing the image data with the pre-stored reference image data, a pre-diagnosis is further included, which comprises: diagnosing the imaging clarity and field of view of the image data; if the diagnosis is passed, the subsequent comparison process is continued; if the imaging clarity does not meet the requirements, an alarm is issued or automatic re-grabbing is performed; if the field of view does not match the standard field of view, the camera parameters are adjusted or a maintenance alarm is issued.

[0007] In an embodiment, the acquisition of the image data of the passageway and work area in the battery swap station comprises: real-time grabbing is realized by calling the software development kit of the multi-camera opposite arrangement; if the grabbing is successful, the image data is transmitted to the diagnosis step; if the grabbing fails, the fault code is captured and the fault reason is analyzed based on the fault code type.

[0008] In an embodiment, diagnosing the imaging clarity of the image data comprises: using at least one of the algorithms of gray variance, gray difference square sum variance, Roberts gradient and / or Laplacian gradient, and evaluating the image quality based on a preset clarity threshold value; if the image quality is greater than or equal to the clarity threshold value, determining that it is clear, and transmitting the image data to subsequent diagnosis or processing steps; if the image quality is less than the clarity threshold value, determining that it is blurred, adjusting the camera exposure time or gain, or re-shooting with optimized settings.

[0009] In an embodiment, diagnosing the field of view of the image data comprises: detecting feature points in a preset area in combination with a target detection algorithm, and determining whether the current field of view matches the standard field of view through an image classification algorithm, wherein the feature points include preset color pattern feature points; if matched, transmitting the image data to obtain a target detection result based on comparison between the image data and a pre-stored vehicle-free and personnel-free reference image; if not matched, adjusting the camera angle or focal length based on the feature point offset, or switching to a backup camera to correct the field of view, or issuing a maintenance alert.

[0010] In an embodiment, the target detection algorithm comprises YOLO, R-CNN or DETR; The image classification algorithm comprises VGG or ResNet.

[0011] In an embodiment, the environmental feature information comprises local background feature points arranged in the channel, and the local background feature points are pre-selected representative features in the battery replacement channel; when judging the existence of vehicles, the target detection algorithm is used to analyze whether the local background feature points are blocked, and the image classification result is combined to determine whether a vehicle exists, so as to improve the recognition accuracy.

[0012] The second aspect of the present application provides a visual judgment system, comprising: an image acquisition module for acquiring image data of a channel and a working area in a battery replacement station in real time; a comparison module for obtaining a target detection result based on comparison between the image data and a pre-stored vehicle-free and personnel-free reference image; a judgment module for judging whether a vehicle and / or personnel exist based on the target detection result and in combination with environmental feature information; a first control module for controlling a roller shutter door to remain open and issuing a safety prompt if a vehicle and / or personnel exist; a second control module for controlling the roller shutter door to close if a vehicle and / or personnel do not exist.

[0013] A third aspect of the present invention provides an electronic device, comprising: a memory for storing instructions executed by one or more processors of the electronic device, and a processor, one of the processors of the electronic device, for the aforementioned visual determination method.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the above-described visual determination method.

[0015] The advantages of this invention over the prior art are as follows: The visual judgment method provided in this application compares real-time acquired image data with preset reference image data to obtain target detection results as the first judgment criterion, and combines environmental feature information as the second judgment criterion. This dual judgment enables accurate determination of the presence of vehicles and / or personnel within the battery swapping station, providing a precise basis for determining whether the roller shutter door remains open or closed. Its rapid door control response in cold weather not only ensures the stability of equipment operation but also significantly reduces energy consumption, achieving both energy saving and high efficiency, making it suitable for practical applications in modern intelligent battery swapping stations. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, 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 the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 According to an embodiment of the present invention, a flowchart of a visual determination method is shown.

[0018] Figure 2 According to an embodiment of the present invention, a schematic diagram of a pre-diagnosis process is shown.

[0019] Figure 3 According to an embodiment of the present invention, a schematic diagram of a process for acquiring image data of passageways and work areas within a battery swapping station is shown.

[0020] Figure 4 According to an embodiment of the present invention, a schematic flowchart for diagnosing the imaging sharpness of image data is shown.

[0021] Figure 5 According to an embodiment of the present invention, a schematic flowchart of a diagnostic image data field of view is shown.

[0022] Figure 6 According to an embodiment of the present invention, a schematic diagram of a visual determination system is shown.

[0023] Figure 7 According to an embodiment of the present invention, a schematic diagram of the structure of an electronic device is shown.

[0024] Figure 8 According to an embodiment of the present invention, a schematic diagram of the structure of a computer-readable storage medium is shown. Detailed Implementation

[0025] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed herein. The present invention can also be implemented or applied through other different specific embodiments, and various details in the present invention can be modified or changed according to different viewpoints and application systems without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0026] The present invention will now be described in detail with reference to the accompanying drawings, so that those skilled in the art can readily implement it. The present invention can be embodied in many different forms and is not limited to the embodiments described herein.

[0027] In the representation of this invention, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics represented in connection with that embodiment or example, which are included in at least one embodiment or example of the invention. Furthermore, the specific features, structures, materials, or characteristics represented may be combined in any suitable manner in one or more embodiments or examples. Moreover, those skilled in the art can combine and integrate different embodiments or examples represented in this invention, as well as features of different embodiments or examples, without contradiction.

[0028] Furthermore, the terms "first" and "second" are used for illustrative purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the representation of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0029] To clearly illustrate the present invention, components unrelated to the description are omitted, and the same or similar constituent elements throughout the specification are given the same reference numerals.

[0030] Throughout this specification, when it is said that a device is "connected" to another device, this includes not only "direct connection" but also "indirect connection" by placing other components in between. Furthermore, when it is said that a device "comprises" a certain constituent element, unless otherwise stated otherwise, this does not exclude other constituent elements, but rather implies that other constituent elements may be included.

[0031] When we say that a device is "above" another device, this can mean that it is directly above the other device, or it can mean that other devices are present in between. Conversely, when we say that a device is "directly" "above" another device, there are no other devices present in between.

[0032] Although the terms first, second, etc., are used in some instances herein to refer to various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. For example, first interface and second interface, etc., are used. Furthermore, as used herein, the singular forms “a,” “an,” and “the” are intended to also include the plural forms unless the context indicates otherwise. It should be further understood that the terms “comprising,” “including,” indicate the presence of features, steps, operations, elements, components, items, kinds, and / or groups, but do not exclude the presence, occurrence, or addition of one or more other features, steps, operations, elements, components, items, kinds, and / or groups. The terms “or” and “and / or” as used herein are interpreted as inclusive, or mean any one or any combination thereof. Thus, “A, B, or C” or “A, B, and / or C” means “any one of: A; B; C; A and B; A and C; B and C; A, B, and C.” Exceptions to this definition will only occur if the combination of elements, functions, steps, or operations is inherently mutually exclusive in some way.

[0033] The technical terms used herein are for reference only to specific embodiments and are not intended to limit the invention. The singular form used herein includes the plural form unless the statement explicitly indicates otherwise. The word "comprising" as used in this specification means to specify a particular characteristic, region, integer, step, operation, element, and / or component, and does not exclude the presence or addition of other characteristics, regions, integers, steps, operations, elements, and / or components.

[0034] Although not explicitly defined, all terms, including technical and scientific terms used herein, shall have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Terms defined in commonly used dictionaries shall be further interpreted as having a meaning consistent with relevant technical literature and the content of this present instruction, and shall not be over-interpreted as having an ideal or overly formulaic meaning unless otherwise defined.

[0035] The visual judgment method proposed in this invention achieves safe, automated, and efficient operation of battery swapping stations through real-time image acquisition, accurate target detection, and intelligent linkage control. Its rapid gating response in cold weather not only ensures the stability of equipment operation but also significantly reduces energy consumption, achieving the dual goals of energy saving and high efficiency. It is suitable for practical application scenarios of modern intelligent battery swapping stations.

[0036] like Figure 1 As shown, a visual determination method includes: Step 110: Real-time acquisition of image data of the passageway and work area within the battery swapping station; specifically, real-time image information of the designated area (passageway and work area) within the battery swapping station is obtained through cameras or other image acquisition devices to provide a data basis for subsequent visual analysis and judgment.

[0037] Step 120: Based on the comparison between the image data and the pre-stored reference image data, obtain the target detection result; it can be understood that the image data is compared with the pre-stored "no vehicles and people" reference image data to detect whether the image contains abnormal targets (such as vehicles or people), generate the target detection result, and provide a basis for subsequent judgment.

[0038] Step 130: Based on the target detection results and combined with environmental feature information, determine whether there are vehicles and / or personnel. It is understood that by using the detection results in step 120 combined with environmental feature information, the accuracy of the judgment of whether there are vehicles or personnel in the battery swapping station can be improved.

[0039] Step 140: If a vehicle and / or personnel are present, the roller shutter door will remain open and a safety warning will be issued. It is understood that when a vehicle or personnel is detected, the roller shutter door will remain open to ensure safe passage, and sound, light or other forms of warning will be issued to remind relevant personnel to pay attention to safety and prevent accidents from happening.

[0040] Step 150: If no vehicles and / or personnel are present, the roller shutter door will close automatically. This means that when it is confirmed that there are no vehicles or personnel inside the battery swapping station, the roller shutter door will automatically close to ensure the safety and security of the station. The following will further explain the specific implementation of steps 110 to 150 above: In step 120 above, the pre-stored reference image data is obtained by collecting images of vehicles present and absent at the battery swapping station, covering various environmental conditions, and classifying the images as "with vehicle" and "without vehicle" for use in training the image classification algorithm model. The trained model can accurately distinguish the station's status based on the occlusion of local background feature points. The pre-stored "no vehicle, no personnel" reference images serve as a comparison benchmark for real-time target detection, ensuring the accuracy and reliability of the visual judgment method.

[0041] In step 130 above, the environmental feature information includes local background feature points set within the channel. These local background feature points are pre-selected representative features within the battery swapping channel. When determining the presence of a vehicle, a target detection algorithm is used to analyze whether the local background feature points are occluded, and the presence of the vehicle is determined in conjunction with the image classification results, thereby improving the recognition accuracy. Specifically, when there is a vehicle in the station, the local background feature points will be occluded by the vehicle and thus cannot be recognized; while when there is no vehicle in the station, the local background feature points will appear completely within the camera's field of view and can be effectively recognized.

[0042] The visual judgment method provided in steps 110 to 150 above acquires image data within the battery swapping station in real time, compares the image data with pre-stored reference image data to obtain target detection results, and combines environmental feature information to determine the presence of vehicles or personnel. This intelligently controls the roller shutter door to remain open and issues a safety warning or automatically closes. When step 130 confirms that there are no battery swapping vehicles or personnel activity within the station, a control signal is immediately generated, automatically triggering the roller shutter door closing mechanism. This rapid-response automated design ensures timely closure of the station environment without manual intervention. Utilizing a pre-trained image classification model, based on occlusion analysis of local background feature points, it accurately distinguishes between "vehicle present / no vehicle" states, thereby achieving precise door control operation.

[0043] The visual judgment method provided in this application achieves safe, automated, and efficient operation of the battery swapping station through real-time image acquisition, accurate target detection, and intelligent linkage control. Its rapid gating response in cold weather not only ensures the stability of equipment operation but also significantly reduces energy consumption, achieving the dual goals of energy saving and high efficiency. It is suitable for practical application scenarios of modern intelligent battery swapping stations.

[0044] In some embodiments of this disclosure, the aforementioned step 120, which compares the image data with pre-stored reference image data, further includes pre-diagnosis. Figure 2 A flowchart illustrating a pre-diagnosis process is shown, such as... Figure 2 As shown, preliminary diagnostics include: Step 121: Diagnose the image sharpness and field of view of the image data; understandably, the image data is compared with the pre-stored reference image data to check whether the previously acquired images meet the sharpness and field of view requirements, ensuring that the image quality is suitable for analysis.

[0045] Step 122: If the diagnosis is successful, continue with the subsequent comparison process; it can be understood that a successful diagnosis means that the image data's imaging clarity and field of view meet the preset requirements, ensuring the accuracy of the subsequent comparison results.

[0046] Step 123: If the image clarity does not meet the requirements, issue an alarm or automatically re-capture the image; this is to obtain clear image data. This process ensures that subsequent analysis is based on high-quality images and avoids misjudgments caused by blurriness.

[0047] Step 124: If the field of view does not match the standard field of view, adjust camera parameters or issue a maintenance alarm. Understandably, if the image field of view does not match the standard field of view (e.g., camera angle shift or obstruction), the system will attempt to adjust camera parameters (such as focal length and angle) to restore the correct field of view; if it cannot adjust automatically, a maintenance alarm will be issued, prompting manual inspection or repair of the camera equipment. The following will further explain the specific implementation of steps 121 to 124 above: The pre-diagnostic method provided in steps 121 and 124 above can determine whether the image data meets quality requirements before comparing it with reference image data, ensuring the accuracy of subsequent comparison results. If the image data fails the diagnosis, alarms, re-capture, or camera parameter adjustments are used to resolve issues such as image blurring or field-of-view mismatch, ensuring stable system operation in complex environments and reducing misjudgments caused by image quality problems. The clarity and field-of-view diagnostic mechanism ensures the accuracy of target detection, avoiding misjudgments of vehicle or personnel presence due to low-quality images, thereby ensuring the correct execution of roller shutter door control and safety prompts and reducing safety risks.

[0048] In some embodiments of this disclosure, Figure 3 The flowchart illustrating the process of collecting image data of the passageway and work area within the battery swapping station in step 110 above is shown below. Figure 3 As shown, the image data collected from the passageways and work areas within the battery swapping station includes... Step 111: Real-time image capture is achieved by utilizing a software development kit (SDK) for a multi-camera, cross-beam layout. Specifically, multiple high-definition cameras are deployed in the battery swapping station's passageways and work areas (such as the battery swapping platform and vehicle access routes) in a cross-beam configuration. For example, cameras can be installed on both sides of the passageway or above and to the side of the work area to ensure coverage of the target area from different angles (such as front, side, and overhead views). This layout effectively captures the dynamics of vehicles or personnel in different positions, reducing blind spots caused by obstructions or a single perspective. Synchronous image capture by multiple cameras is achieved by utilizing the SDK provided by the camera manufacturer. The SDK supports interfaces including triggering capture, setting resolution, and adjusting frame rate. For example, the system can be configured to capture images at 1080p resolution and 5 frames per second to meet the real-time monitoring needs of the battery swapping station.

[0049] Step 112: If the image capture is successful, the image data is transmitted to the diagnostic step. Specifically, if the image capture is successful, the acquired image data is sent to the subsequent image diagnostic step (Step 120) via a data transmission module (e.g., via network or local storage). The transmission process must ensure data integrity and real-time performance to avoid image loss or delay. Image data is typically transmitted in a standard format (e.g., JPEG or PNG) with accompanying metadata (e.g., timestamp, camera number) to facilitate subsequent diagnosis and analysis. The transmission system may use wired (e.g., Ethernet) or wireless (e.g., Wi-Fi) methods, depending on the hardware configuration of the battery swapping station.

[0050] Step 113: If the capture fails, capture the fault code and analyze the cause of the failure based on the fault code type. Specifically, if the capture fails, the system will capture the fault code and analyze the cause of the failure based on the fault code type (such as camera hardware failure, communication interruption, insufficient light, etc.). Based on the analysis results, the system may trigger automatic remedial measures (such as restarting the camera, adjusting exposure parameters) or issue a maintenance alarm to prompt manual intervention. The fault code is usually generated by the SDK or the camera system and contains specific error information (such as "lens obstruction" or "network connection failure"). The system quickly locates the problem and records the log through preset fault handling logic, providing a basis for subsequent maintenance. The following will further explain the specific implementation of steps 111 to 113 above: In step 111 mentioned above, the capture covers various environmental conditions (such as daytime, nighttime, and rainy weather). By automatically adjusting exposure, gain, or white balance, the clarity of the image is ensured under different lighting conditions. The SDK also supports infrared or low-light imaging modes to enhance the capture effect in nighttime or low-light environments. The SDK synchronizes multiple cameras through timestamps to ensure that the image data at the same time is consistent, which facilitates subsequent multi-view analysis. The captured images are stored in a high compression ratio format (such as JPEG) to reduce storage and transmission pressure while retaining key details (such as vehicle outlines and background feature points). The capture frequency is dynamically adjusted according to the operational needs of the battery swapping station. For example, the frame rate is increased during peak hours to capture fast-moving targets, and the frame rate is reduced during off-peak hours to save resources.

[0051] In step 112 mentioned above, the successfully captured image data is quickly sent to the image diagnostic step (step 120) via the data transmission module. The transmission module is typically based on the network architecture of the battery swapping station, for example, connecting to a local server via Ethernet or uploading to a cloud processing platform via Wi-Fi. To meet real-time monitoring requirements, the system optimizes transmission latency, for example, by transmitting image streams via low-latency protocols (such as RTSP or WebRTC), ensuring that the latency from capture to diagnosis is controlled within milliseconds. Each image is accompanied by metadata, including capture time, camera ID, location information, etc. For example, an image might be labeled "2025-08-03 09:00:01, Camera_01, Channel Entrance". This metadata facilitates the system in tracing the image source and assists in subsequent diagnosis and target detection.

[0052] The aforementioned steps 111 to 113, through the arrangement of multiple cameras, reliable data transmission, and fault handling, have enabled the comprehensive and stable acquisition of image data of the passage and work area within the battery swapping station, laying the foundation for subsequent diagnosis (steps 121-122) and target detection (steps 120 and 130).

[0053] In some embodiments of this disclosure, Figure 4 This diagram illustrates the process of improving the imaging sharpness of the diagnostic image data in step 121 above; as shown. Figure 4 As shown, the process for diagnosing the imaging sharpness of image data includes: Step 210: Employ at least one algorithm selected from grayscale variance, grayscale difference sum of squares variance, Roberts gradient, and / or Laplacian gradient, and evaluate image quality based on a preset sharpness threshold. Specifically, apply the aforementioned algorithms to the image data transmitted in step 112 (generated by multi-camera capture in step 111), calculate a sharpness score, and compare it with a preset sharpness threshold. The threshold is pre-calibrated based on the battery swapping station scenario (e.g., lighting conditions, target detection requirements), for example, set to a specific value of grayscale variance or the minimum threshold of the Laplacian gradient.

[0054] Step 220: If the image quality is greater than or equal to the sharpness threshold, it is determined to be sharp, and the image data is transmitted to subsequent diagnostic or processing steps. Specifically, the sharp image is sent to subsequent diagnostic steps (such as the field of view diagnosis in step 122) or target detection steps (step 120) via a data transmission module (similar to the network or local storage in step 112). Metadata (such as timestamps and camera IDs) is included in the transmission to ensure data traceability.

[0055] Step 230: If the image quality is below the sharpness threshold, it is determined to be blurry. The camera exposure time or gain is adjusted, or the image is re-captured with optimized settings. Specifically, adjusting camera parameters: The system calls the SDK from step 111 to adjust the camera exposure time (e.g., extending the exposure to increase brightness) or gain (e.g., increasing signal amplification to enhance details) to attempt to improve image quality. Re-capture: If parameter adjustment is ineffective, the system triggers a re-capture (calling the multi-camera capture function from step 111) to acquire a new image with optimized settings. The new settings may be based on the fault analysis results from step 113 (e.g., adding supplementary lighting when light is insufficient). Alarm mechanism: If a clear image cannot be obtained after multiple attempts, the system issues an alarm (similar to step 121), notifying maintenance personnel to check for environmental or hardware problems (e.g., lens smudges, light source failure). The following will further explain the specific implementation of steps 210 to 230 above: In step 210 above, grayscale variance: by calculating the variance of the grayscale values ​​of image pixels, the overall contrast and detail richness of the image are evaluated. The larger the variance, the richer the image details, which usually indicates that it is clearer.

[0056] Gray-level difference sum of squares variance: The variance is calculated based on the sum of squares of the gray-level differences between adjacent pixels, highlighting local contrast changes in the image and suitable for detecting edge sharpness.

[0057] Roberts gradient: Calculates the gradient magnitude of an image using the Roberts operator, emphasizing edge strength, and is suitable for quickly detecting whether an image has clear boundary features.

[0058] Laplacian gradient: Detects high-frequency components of an image using the Laplacian operator, reflecting the sharpness of image details, and is particularly suitable for evaluating overall sharpness.

[0059] Steps 210 to 230 described above evaluate image sharpness using algorithms such as grayscale variance, Roberts gradient, and Laplacian gradient (step 210), ensuring that clear images are transmitted to subsequent steps (step 220), and processing blurry images by adjusting camera parameters or re-capturing (step 230). These steps refine the image diagnostic process and seamlessly connect with the acquisition steps 111-113 and the diagnostic steps 121-122, jointly supporting the intelligent monitoring and management of battery swapping stations. In cold weather, this method significantly improves thermal insulation performance and reduces energy consumption through fast and accurate image processing and gating response, achieving safe, energy-saving, and efficient operation, making it suitable for modern intelligent battery swapping station scenarios.

[0060] In some embodiments of this disclosure, Figure 5 The flowchart illustrating the visual field of the diagnostic image data in step 121 above is shown, as follows: Figure 5 As shown, the process of visual field analysis for diagnostic image data includes: Step 310: Combine the target detection algorithm to detect feature points within the preset area, and use the image classification algorithm to determine whether the current field of view matches the standard field of view. The feature points include preset color pattern feature points. Specifically, the target detection algorithm includes YOLO, R-CNN or DETR; the image classification algorithm includes VGG or ResNet.

[0061] Among them, the target detection algorithm is: YOLO: Quickly detects feature point bounding boxes and locations, suitable for real-time scenarios.

[0062] R-CNN: Improves feature point detection accuracy through region proposal networks, making it suitable for complex environments.

[0063] DETR: Based on the Transformer architecture, it is suitable for multi-target scenarios and efficiently detects multiple feature points.

[0064] Image classification algorithms: VGG: Extracts feature point distribution characteristics through deep convolutional networks to determine field-of-view matching.

[0065] ResNet: Improves classification accuracy by using residual networks and adapts to changes in lighting or partial occlusion.

[0066] Step 320: If a match is found, the image data is transmitted for comparison with pre-stored reference images showing no vehicles or people to obtain a target detection result. Specifically, the matched image is sent to the target detection step (step 120) via a data transmission module (similar to steps 112 and 220) for comparison with pre-stored "no vehicles, no people" reference images to generate a target detection result (e.g., whether a vehicle or person exists within the station). Metadata (such as timestamps, camera IDs, and match confidence scores) is transmitted for easy tracking and analysis.

[0067] Step 330: If a mismatch occurs, adjust the camera angle or focal length based on the feature point offset, or switch to a backup camera to correct the field of view, or issue a maintenance alert. Specifically, adjust camera parameters: Based on the feature point offset (the pixel offset vector calculated by the object detection algorithm), adjust the camera angle (e.g., gimbal rotation) or focal length (e.g., zoom in / out) using the SDK in step 111 to correct the field of view to the standard field of view. Switch to backup camera: If the adjustment is ineffective (e.g., the main camera is obstructed, hardware failure, or IP configuration error), the system switches to a backup camera (utilizing the redundancy of the multi-camera setup in step 111) and recaptures the image to restore the correct field of view. Maintenance alert: If a mismatch still occurs after multiple attempts (e.g., feature points are continuously missing or the camera ID is incorrect), the system issues a maintenance alert (similar to steps 121 and 230), notifying maintenance personnel to check the camera status, installation angle, or IP configuration (e.g., "Camera_01 field of view offset, it is recommended to check the installation angle or IP configuration").

[0068] The following will further explain the specific implementation of steps 310 to 330 above: In step 310 above, feature point detection: a target detection algorithm (such as YOLO, R-CNN or DETR) is used to detect unique feature points in a preset area. These feature points are preset color pattern feature points in the passage or work area of ​​the battery swapping station, such as red squares on the wall, green triangles on the floor, high-contrast markings on the equipment, etc. These feature points have fixed positions and distributions in the standard field of view, which are used to verify whether the camera's field of view accurately covers the preset area (such as the passage entrance, battery swapping platform).

[0069] Field of view matching judgment: By using an image classification algorithm (such as VGG or ResNet), compare whether the feature point distribution of the current image is consistent with the pre-stored standard field of view (i.e., the feature point distribution in the reference image with "no vehicles and no people"); the classification algorithm outputs the matching result (such as "match" or "no match") to determine whether the current camera's field of view correctly covers the key area.

[0070] Preset camera verification: To resolve the issue of inconsistent camera IP configurations, the system verifies whether the current image comes from a preset camera by detecting feature points when calling the snapshot function (step 111). Each camera corresponds to a unique standard field-of-view template (containing a specific feature point distribution and classification labels). If the detected feature points do not match the template, it indicates that the wrong camera may have been called (IP configuration error).

[0071] Angle deviation detection: The camera installation angle is determined by the offset of the feature point position (calculated by the target detection algorithm). If the offset exceeds the tolerance range (e.g., the feature point deviates from the standard position by more than 10 pixels), it indicates that there is an installation angle problem.

[0072] Model training: Standard images of the field of view of each camera within the station are collected in advance (such as markers on both sides of the channel and positioning patterns of the work area), and a classification label is assigned to each camera's field of view (such as "Camera_01_Channel Entrance"). This data is used to train an image classification model (such as ResNet). After training, the model infers whether the field of view of the current image matches the preset label.

[0073] In step 320 above, during transmission, the metadata includes the camera ID and matching tag to ensure that the image comes from the preset camera and avoid erroneous data transmission caused by IP configuration confusion. Recording the metadata, feature point detection results, and classification confidence of the matched image (e.g., "Camera_01, matching confidence 0.98") facilitates subsequent troubleshooting of installation or configuration issues.

[0074] In step 330 above, angle deviation correction: offset analysis is accurate to the pixel level and converted into gimbal rotation angle or focal length adjustment parameters to ensure that the corrected field of view is consistent with the standard field of view.

[0075] IP Configuration Check: If the feature point distribution indicates that the image comes from the wrong camera, the system records the wrong camera ID, triggers an IP configuration check, and prompts maintenance personnel to reassign or confirm network settings.

[0076] Environmental interference handling: If the mismatch is caused by obstruction or dirt (such as feature points being covered by obstacles), the alarm will clearly indicate the problem type (such as "Camera_02 Lens obstruction"), which will facilitate quick maintenance.

[0077] Application scenario: In severe weather (such as heavy snow covering the lens) or during the installation and debugging phase (such as the angle deviation of a newly installed camera), step 330 can quickly restore the field of view through parameter adjustment, switching to a backup camera, or maintenance alarm.

[0078] Steps 310 to 330 detect feature points of a preset color pattern using target detection algorithms (YOLO, R-CNN, DETR) and combine them with image classification algorithms (VGG, ResNet) to determine field-of-view matching (step 310). Matching images are transmitted to target detection (step 320), while non-matching images are handled through camera parameter adjustment, backup camera switching, or maintenance alarm processing (step 330). This process avoids issues related to camera installation angle deviation and IP configuration confusion. It seamlessly integrates with steps 111-113 (acquisition), 210-230 (clarity diagnosis), and 121-122 (diagnostic processing), supporting intelligent monitoring and management of battery swapping stations. In cold weather, this method improves thermal insulation performance and reduces energy consumption through fast and accurate image processing and gating response, achieving safe, energy-saving, and efficient operation, making it suitable for modern intelligent battery swapping station scenarios.

[0079] In a preferred embodiment, in order to improve the operational reliability and judgment accuracy of the automatic gate control system of the battery swapping station under complex environmental conditions, the present invention integrates multi-camera beam acquisition, field of view diagnosis and correction, and vehicle presence detection based on local background feature point occlusion and forms a collaborative working mechanism.

[0080] Specifically, firstly, multiple sets of through-beam cameras deployed in the battery swapping station passage and work area acquire original images from multiple perspectives through synchronous triggering. During the acquisition process, imaging success diagnosis and fault code analysis are performed, and hardware status detection and communication status detection are conducted on the acquisition results. Image data with imaging failure, communication interruption or image abnormality are discarded, and only normal image data that meets the imaging quality requirements are retained for subsequent processing.

[0081] Secondly, the system analyzes the transmitted image data based on a preset feature point detection algorithm (such as YOLO, R-CNN, or DETR), extracts preset color pattern feature points from the image, and combines the classification results of an image classification algorithm (such as VGG or ResNet) to determine whether the current camera's field of view matches the standard reference field of view. If the field of view does not match, the system calls the pan-tilt control command to adjust the camera's mounting angle or focal length, or switches to a backup camera, and re-executes the capture until the current field of view matches the standard reference field of view. This step ensures that local background feature points are fully presented in the image imaging area, providing high-quality and highly stable basic data for vehicle presence detection.

[0082] Secondly, during the vehicle identification phase, the system analyzes whether local background feature points are obscured by vehicles in the image confirmed by field-of-view matching, and combines this with the classification results of the entire image to comprehensively determine whether a vehicle exists within the battery swapping station. Furthermore, when the detection results indicate that some local background feature points are missing for an extended period, and the image classification result shows a "no vehicle" status, the system will trigger a reverse call to the field-of-view diagnostic process to re-verify the camera's installation position, focal length setting, and lens cleanliness. This forms a closed-loop operating mechanism combining forward detection and reverse correction, reducing misjudgments and missed judgments caused by camera equipment offset, focal length errors, or lens contamination.

[0083] Through the above implementation methods, this invention achieves a close integration of multi-camera joint acquisition, automated field-of-view protection, and vehicle presence determination based on background feature point occlusion in a real operating environment. This not only improves the accuracy of vehicle detection and gate control linkage response but also enhances the system's adaptability to equipment offset, environmental changes, and lighting interference. In some embodiments of this disclosure, Figure 6 A schematic diagram of a vision-based decision-making system is shown, such as... Figure 6 As shown, a visual determination system includes: Image acquisition module 501: Used to acquire image data of the passageway and work area within the battery swapping station; The comparison module 502 compares the image data with pre-stored reference images without vehicles and people to obtain target detection results; The judgment module 503 is used to determine whether there are vehicles and / or personnel based on the target detection results and combined with environmental feature information; The first control module 504, if there are vehicles and / or personnel, will link the control to keep the roller shutter door open and issue a safety warning. The second control module 505, if there are no vehicles and / or personnel, will control the roller shutter door to close.

[0084] Those skilled in the art will understand that various aspects of this disclosure can be implemented as a system, method, or program product. Therefore, various aspects of this disclosure can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software aspects, collectively referred to herein as a "circuit," "module," or "platform."

[0085] Specifically, Figure 7 A schematic diagram of the structure of an electronic device is shown according to an embodiment of this disclosure. Referring below... Figure 7 To describe an electronic device 600 according to such an embodiment of the present disclosure. Figure 7 The electronic device 600 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0086] like Figure 7 As shown, the electronic device 600 is presented in the form of a general-purpose computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including storage unit 620 and processing unit 610), a display unit 640, etc.

[0087] The storage unit stores program code, which can be executed by the processing unit 610, causing the processing unit 610 to perform steps according to various exemplary embodiments of this disclosure. For example, the processing unit 610 can perform actions such as... Figure 1 The steps of the visual determination method shown are as follows.

[0088] Storage unit 620 may include a readable medium in the form of a volatile storage unit, such as random access memory (RAM) 6201 and / or cache storage unit 6202, and may further include a read-only storage unit (ROM) 6203.

[0089] Storage unit 620 may also include a program / utility 6204 having a set (at least one) program module 6205, such program module 6205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0090] Bus 630 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0091] Electronic device 600 can also communicate with one or more external devices 700 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 650. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 660. Network adapter 660 can communicate with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0092] This disclosure also provides a computer-readable storage medium for storing a program, which, when executed, implements the steps of a visual determination method. In some possible implementations, various aspects of this disclosure can also be implemented as a program product including program code that, when run on a terminal device, causes the terminal device to perform the steps described in the foregoing document generation method section of this specification according to various exemplary embodiments of this disclosure.

[0093] Specifically, Figure 8 According to an embodiment of this disclosure, a schematic diagram of the structure of a computer-readable storage medium is shown. For example... Figure 8 As shown, a program product 800 for implementing the above-described visual determination method according to an embodiment of the present disclosure is described. It may employ a portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, system, or device.

[0094] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or a combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0095] Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium may also be any readable medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, system, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, or any suitable combination thereof.

[0096] Program code for executing specific implementations of the visual determination method provided in the foregoing embodiments of this disclosure can be written in any combination of one or more programming languages. Programming languages ​​include object-oriented programming languages—such as Java, C++, etc.—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0097] In summary, the proposed visual judgment method, through real-time image acquisition, accurate target detection, and intelligent linkage control, achieves safe, automated, and efficient operation of the battery swapping station. Its rapid gating response in cold weather not only ensures the stability of equipment operation but also significantly reduces energy consumption, achieving the dual goals of energy saving and high efficiency. It is suitable for practical application scenarios of modern intelligent battery swapping stations.

[0098] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A visual determination method, characterized in that, include: Real-time acquisition of image data of the passageways and work areas within the battery swapping station; The target detection result is obtained by comparing the image data with the pre-stored reference image data; Based on the target detection results and combined with environmental feature information, determine whether there are vehicles and / or personnel. If a vehicle and / or person is present, the linkage control will keep the roller shutter door open and issue a safety warning. If there are no vehicles and / or personnel, the roller shutter door will close automatically.

2. The visual determination method as described in claim 1, characterized in that: Before comparing the image data with pre-stored reference image data, a preliminary diagnosis is also included, which includes: Diagnose the imaging sharpness and field of view of the image data; If the diagnosis is successful, proceed with the subsequent comparison process; If the image clarity does not meet the requirements, an alarm will be issued or the image will be automatically re-captured; If the field of view does not match the standard field of view, adjust the camera parameters or issue a maintenance alarm.

3. The visual determination method as described in claim 2, characterized in that: Diagnosing the image sharpness of the image data includes: Image quality is evaluated using at least one of the following algorithms: gray-level variance, gray-level difference sum of squares variance, Roberts gradient and / or Laplacian gradient, based on a preset sharpness threshold. If the image quality is greater than or equal to the sharpness threshold, it is determined to be sharp, and the image data is transmitted to subsequent diagnostic or processing steps. If the image quality is less than the sharpness threshold, it is determined to be blurry. The camera exposure time or gain is adjusted, or the image is retaken with optimized settings.

4. The visual determination method as described in claim 2, characterized in that: The diagnostic field of view for the image data includes: The algorithm combines a target detection algorithm to detect feature points within a preset area and uses an image classification algorithm to determine whether the current field of view matches the standard field of view. The feature points include preset color pattern feature points. If a match is found, image data is transmitted to perform a comparison between the image data and a pre-stored reference image without vehicles or people to obtain a target detection result. If there is a mismatch, adjust the camera angle or focal length based on the feature point offset, switch to a backup camera to correct the field of view, or issue a maintenance alarm.

5. The visual determination method as described in claim 4, characterized in that: The target detection algorithm includes YOLO, R-CNN, or DETR; The image classification algorithm includes VGG or ResNet.

6. The visual determination method as described in claim 1, characterized in that: The image data collected from the passageways and work areas within the battery swapping station includes: Real-time image capture is achieved by calling a software development kit that uses multiple cameras in a cross-beam configuration. If the capture is successful, the image data will be transmitted to the diagnostic steps. If the capture fails, the fault code is captured and the cause of the fault is analyzed based on the fault code type.

7. The visual determination method as described in claim 1, characterized in that: The environmental feature information includes local background feature points set within the channel, and the local background feature points are pre-selected representative features within the battery swapping channel; When determining the presence of a vehicle, the local background feature points are analyzed using a target detection algorithm to determine whether they are occluded, and the presence of the vehicle is determined in conjunction with the image classification results, thereby improving the recognition accuracy.

8. A visual judgment system, characterized in that, include: Image acquisition module: Used to acquire image data of the passageway and work area within the battery swapping station in real time; The comparison module is used to compare the image data with pre-stored reference images without vehicles and people to obtain target detection results; The judgment module is used to determine whether there are vehicles and / or personnel based on the target detection results and combined with environmental feature information; The first control module, if a vehicle and / or personnel are present, will link the control to keep the roller shutter door open and issue a safety warning. The second control module, if there are no vehicles and / or personnel, will activate the linkage control to close the roller shutter door.

9. An electronic device, characterized in that, include: A memory for storing instructions executed by one or more processors of an electronic device, and a processor, one of the processors of the electronic device, for executing the visual determination method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the visual determination method as described in any one of claims 1 to 7.

Images