Multi-situation photographing edge analysis method

Abstract

The invention discloses a multi-situation photographing edge analysis method. According to the method, an environment map is constructed in real time through an improved SLAM algorithm, the position of the robot is positioned, and all-around environment perception is achieved through multiple types of sensors. After the robot arrives at an inspection point, the angle and the height of a camera are adjusted through a mechanical arm to collect images, and preprocessing operations such as denoising and contrast enhancement are carried out on the images. And then instrument data in the image is identified by using a deep learning model and an OCR technology, and preliminary data analysis is carried out on edge computing equipment. Important data are uploaded to a cloud end through a wireless network for further processing. According to the invention, the inspection efficiency and safety are improved, the data accuracy and reliability are improved, and the data management and real-time performance are optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent inspection robots, and in particular to a multi-stance photography edge analysis method. Background Art

[0002] In the current field of electrical equipment maintenance and monitoring, especially in high-voltage dangerous environments such as hotel power distribution rooms, traditional inspection methods mainly rely on manual inspections. This method has many limitations and risks:

[0003] First, manual inspections require professional technicians to regularly enter the distribution room and manually record parameters such as voltage and current on various instruments. This method is not only time-consuming and laborious, but also has certain safety hazards, such as the risk of electric shock and the possibility of accidents. Secondly, although some places have deployed fixed surveillance cameras or sensor systems for real-time monitoring of the status of electrical equipment, these systems are usually installed in fixed positions, cannot flexibly adjust the viewing angle, and are difficult to cover all key areas. In addition, although the existing remote monitoring systems can achieve a certain degree of data collection and transmission, they often lack intelligent analysis capabilities and cannot automatically identify specific numerical information in the image. Human intervention is still required for data analysis. In terms of technical means, although there are autonomous navigation inspection technologies, data return edge computing technologies, and image recognition algorithm technologies, there are still some problems with existing technologies, such as the inability to support network-free environments, and sometimes there is an impact on autonomous cruising in high magnetic fields or strong pulse work locations.

[0004] Therefore, the existing technology cannot simultaneously meet the requirements for improving safety, efficiency and accuracy during the inspection of electrical equipment in high-voltage dangerous environments. Summary of the invention

[0005] Based on this, an embodiment of the present application provides a multi-situation photographing edge analysis method, which achieves safer and more effective electrical equipment inspection by performing multi-situation photographing edge analysis in a high-voltage dangerous environment.

[0006] In a first aspect, a multi-stance photography edge analysis method is provided, the method comprising:

[0007] The improved SLAM algorithm is used to construct an environmental map in real time and locate the position of the detection robot, and the multi-type sensors set on the detection robot are used to perform all-round environmental perception of the environment to be detected; wherein the multi-type sensors include at least a laser radar, an ultrasonic sensor and a camera;

[0008] When the inspection robot arrives at the designated inspection point, it performs image acquisition and preprocessing. Image acquisition includes adjusting the camera shooting angle and height through the preset parameters of the robot arm, and preprocessing includes denoising, contrast enhancement, and cropping the acquired images.

[0009] Use deep learning models to extract features from preprocessed images and combine them with OCR technology to identify numbers and symbols on instruments in the images;

[0010] The identified data is processed and preliminarily analyzed on the edge computing device, and the important data analyzed are uploaded to the cloud via the wireless network.

[0011] Optionally, the step of using the improved SLAM algorithm to construct an environment map in real time and locate the position of the detection robot also includes:

[0012] Combined with A* or Dijkstra path planning algorithms, the optimal path from the current position to the next inspection point is planned for the inspection robot based on the real-time constructed environment map, and the path is dynamically adjusted when encountering obstacles.

[0013] Optionally, the steps of image acquisition and preprocessing further include:

[0014] The multi-degree-of-freedom robotic arm dynamically adjusts the camera's shooting angle and height based on the instrument's position and posture detected in real time to ensure that the instrument is fully displayed in the center area of ​​the captured image.

[0015] Optionally, the step of extracting features from the preprocessed image using a deep learning model further includes:

[0016] A convolutional neural network is used to perform multi-layer feature extraction on the image to extract the key feature areas in the image, including the outline of the dashboard, the position of the pointer, and the shape features of the digital characters.

[0017] Optionally, the step of using OCR technology to recognize numbers and symbols on instruments in the image further includes:

[0018] Combined with the context analysis algorithm, the OCR recognition results are corrected. By comparing the rationality of historical data and current recognition results, recognition errors caused by image quality or character ambiguity can be corrected.

[0019] Optionally, the step of performing data processing and preliminary data analysis on the identified data on the edge computing device further includes:

[0020] The raw data is filtered and compressed on the edge computing device, retaining only the feature data related to the meter readings, and the data is analyzed for anomaly detection to mark abnormal data points that exceed the preset threshold range.

[0021] In a second aspect, a multi-state photographing edge analysis system is provided, the system comprising:

[0022] A positioning detection module is used to construct an environmental map in real time using an improved SLAM algorithm and locate the position of the detection robot, and to perform all-round environmental perception of the environment to be detected through multiple types of sensors arranged on the detection robot; wherein the multiple types of sensors include at least a laser radar, an ultrasonic sensor and a camera;

[0023] The acquisition and processing module is used to acquire and preprocess images after the detection robot reaches the designated inspection point; image acquisition includes adjusting the camera shooting angle and height through the preset parameters of the robot arm, and preprocessing includes denoising, contrast enhancement, and cropping the acquired images;

[0024] The feature extraction module is used to extract features from the preprocessed image using a deep learning model and recognize numbers and symbols on instruments in the image in combination with OCR technology;

[0025] The analysis and processing module is used to process and perform preliminary data analysis on the identified data on the edge computing device, and upload the important data analyzed by the data analysis results to the cloud via the wireless network.

[0026] In a third aspect, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the multi-stance photography edge analysis method described in any one of the first aspects is implemented.

[0027] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the multi-stance photography edge analysis method described in any one of the first aspects is implemented.

[0028] In a fifth aspect, a computer program product is provided, including a computer program / instruction, which, when executed by a processor, implements the multi-stance photography edge analysis method described in any one of the first aspects above.

[0029] The beneficial effects brought by the technical solution provided by the embodiment of the present application include at least:

[0030] (1) Improved safety: The risk of direct contact between personnel and high-voltage equipment is avoided, reducing the possibility of personal injury; the robot can automatically avoid obstacles and ensure safe movement in narrow spaces, further reducing the possibility of accidents.

[0031] (2) Improve work efficiency: The inspection time is greatly shortened and the inspection efficiency is improved. Since the inspection speed is faster, the inspection frequency can be increased and potential problems can be discovered in time.

[0032] (3) Enhanced data accuracy: Image quality and recognition accuracy are ensured through image preprocessing and deep learning models; the automated recognition process reduces errors caused by human factors and improves data reliability.

[0033] (4) Optimize data management: Local edge computing devices enable fast data processing, reduce data transmission time, and improve the system's real-time response capability; selectively upload useful data, reduce network load, and ensure data security and integrity; efficient data management and storage mechanisms facilitate subsequent query and analysis, improving the overall management level. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0035] Figure 1 A flowchart of the steps of a multi-state photo edge analysis method provided in an embodiment of the present application;

[0036] Figure 2 A flowchart provided for an optional embodiment of the present application;

[0037] Figure 3 A flowchart provided for an optional embodiment of the present application;

[0038] Figure 4 A block diagram of a multi-state camera edge analysis system provided in an embodiment of the present application;

[0039] Figure 5 A schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0041] In the description of the present invention, the terms "comprises", "has" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may also include other steps or units that are not explicitly listed but are inherent to these processes, methods, products or apparatuses, or steps or units added based on further optimization schemes conceived by the present invention.

[0042] In the current field of electrical equipment maintenance and monitoring, especially in high-voltage dangerous environments such as hotel distribution rooms, traditional methods mainly rely on manual inspections. This method has many limitations and risks:

[0043] Manual inspection: Traditional electrical equipment inspections are usually conducted by professional technicians, who need to regularly enter the power distribution room and manually record parameters such as voltage and current on various instruments. This method is not only time-consuming and labor-intensive, but also poses certain safety risks.

[0044] Fixed monitoring systems: Some sites have deployed fixed surveillance cameras or sensor systems to monitor the status of electrical equipment in real time. However, these systems are usually installed in fixed locations and cannot flexibly adjust the viewing angle, making it difficult to cover all key areas.

[0045] Remote monitoring and data transmission: Although existing remote monitoring systems can achieve a certain degree of data collection and transmission, they often lack intelligent analysis capabilities and cannot automatically identify specific numerical information in images. Human intervention is still required for data analysis.

[0046] 1. Autonomous navigation inspection technology:

[0047] Sensor system: including LiDAR, ultrasonic sensors, cameras, etc., used to perceive the surrounding environment in real time.

[0048] Navigation algorithm: Based on the SLAM (Simultaneous Localization and Mapping) algorithm, the robot can create a map of the environment and determine its own position.

[0049] Path planning: Through path planning algorithms such as A* and Dijkstra, the robot can find the optimal path from the starting point to the target point and automatically adjust the path when encountering obstacles.

[0050] Obstacle avoidance: Using sensor data and obstacle avoidance algorithms, the robot can detect and avoid obstacles in real time to ensure safe movement.

[0051] 2. Data backhaul edge computing technology:

[0052] Edge device: A small computing device deployed locally that has certain computing power and storage space and can perform some data analysis tasks.

[0053] Data filtering and compression: Filter and compress raw data on edge devices to upload only useful information and reduce transmission volume.

[0054] Real-time processing: Use edge computing capabilities to perform preliminary data processing and analysis, such as anomaly detection and trend prediction.

[0055] Cloud collaboration: Important data and analysis results can be uploaded to the cloud via wireless networks for more in-depth analysis and long-term storage.

[0056] 3. Image recognition algorithm technology:

[0057] Image preprocessing: including denoising, contrast enhancement, cropping and other operations to improve the accuracy of subsequent recognition.

[0058] Feature extraction: Use deep learning models such as convolutional neural networks (CNNs) to extract key features from images.

[0059] Character Recognition: Apply optical character recognition (OCR) technology to convert numbers and symbols in images into readable text data.

[0060] Data analysis: Combine contextual information and historical data to further analyze and verify the recognition results.

[0061] 4. Problems:

[0062] It cannot support an environment without a network. The lack of a network will affect the algorithm's analysis results of the captured pictures. It may sometimes affect autonomous cruising in work locations facing high magnetic fields or strong pulses.

[0063] The invention aims to realize an autonomous mobile camera robot system by accessing a multi-state camera edge analysis method to solve the safety, efficiency and accuracy issues of electrical equipment inspection in high-voltage dangerous environments. Specific goals include:

[0064] Improve safety: Avoid direct human contact with high-voltage equipment and reduce potential safety risks.

[0065] Improve work efficiency: realize automated inspection, quickly complete data collection of multiple instruments, and significantly shorten inspection time.

[0066] Enhanced data accuracy: Using advanced image recognition technology, key parameters such as voltage and current can be accurately extracted to avoid manual reading errors.

[0067] Increased flexibility: The robot can flexibly adjust the shooting angle and height according to actual needs to ensure that every key area can be effectively monitored.

[0068] Optimize data management: Through edge computing and data backhaul technology, efficient data processing and storage can be achieved to facilitate subsequent query and analysis.

[0069] The invention provides a more intelligent, reliable and efficient solution for the inspection of electrical equipment in high-voltage dangerous environments, ensuring the safety and accuracy of the inspection work.

[0070] The invention belongs to the technical field of intelligent inspection robots, and specifically relates to a high-voltage distribution equipment monitoring system based on autonomous navigation and machine vision. Its technical scope covers: power system automation monitoring devices, industrial robot structural design with environmental self-adaptation capabilities, non-contact power parameter identification algorithms, and the development of high-risk environment operation platforms integrating multi-sensor data. This technology is directly applied to the intelligent inspection of unmanned high-risk power sites, integrating robot motion control technology and image recognition and processing technology (including instrument panel digital recognition and indicator light status detection), and is suitable for all-weather automated safety monitoring of 10kV and below distribution equipment in scenarios such as hotels, factories, and substations.

[0071] Please refer to Figure 1 , which shows a flow chart of a multi-state photographing edge analysis method provided by an embodiment of the present application, which may include the following steps:

[0072] S1 uses the improved SLAM algorithm to build an environmental map in real time and locate the position of the detection robot, and uses multiple types of sensors installed on the detection robot to perform all-round environmental perception of the environment to be detected.

[0073] Among them, the multiple types of sensors include at least lidar, ultrasonic sensors and cameras.

[0074] In step S1, the inspection robot perceives the surrounding environment in real time through integrated multi-type sensors (including LiDAR, ultrasonic sensors and cameras). LiDAR is used to measure distance and generate two-dimensional or three-dimensional point cloud data of the environment, ultrasonic sensors are used for close-range obstacle detection, and cameras are used to obtain visual information of the environment. The improved SLAM algorithm combines these sensor data to build an environmental map in real time and accurately locate the robot's position. At the same time, based on the A* or Dijkstra path planning algorithm, the robot can plan the optimal path from the current position to the next inspection point according to the real-time map, and dynamically adjust the path when encountering obstacles, ensuring efficient movement and obstacle avoidance in complex environments.

[0075] S2, when the detection robot arrives at the designated inspection point, it performs image acquisition and preprocessing.

[0076] Among them, image acquisition includes adjusting the camera shooting angle and height through the preset parameters of the robotic arm, and preprocessing includes denoising, contrast enhancement, and cropping operations on the collected images.

[0077] In step S2, after the robot arrives at the designated inspection point, it flexibly adjusts the shooting angle and height of the camera through the multi-degree-of-freedom robotic arm to ensure that the instrument is fully displayed in the central area of ​​the captured image. The robotic arm dynamically adjusts according to preset parameters or the position and posture of the instrument detected in real time to adapt to the installation position and angle of different instruments. After the image acquisition is completed, it enters the preprocessing stage, including denoising, contrast enhancement and cropping of the collected image to improve the image quality and provide a reliable data basis for subsequent image recognition. Through these preprocessing operations, the image quality degradation caused by insufficient light or reflection can be effectively reduced to ensure that the key information in the image is clear and discernible.

[0078] S3 uses a deep learning model to extract features from preprocessed images and combines it with OCR technology to recognize numbers and symbols on instruments in the image.

[0079] In step S3, the preprocessed image is first subjected to multi-layer feature extraction by a convolutional neural network (CNN) to extract key feature areas in the image, including the outline of the dashboard, the position of the pointer, and the shape features of the digital characters. Subsequently, the optical character recognition (OCR) technology is combined to convert the numbers and symbols in the image into readable text. In order to further improve the accuracy of recognition, the system will correct the OCR recognition results in combination with the context analysis algorithm, and correct the recognition errors caused by image quality or character ambiguity by comparing the rationality of historical data and current recognition results.

[0080] S4, performs data processing and preliminary data analysis on the identified data on the edge computing device, and uploads the important data of the data analysis results to the cloud via the wireless network.

[0081] In step S4, the identified data is first processed on the edge computing device. The edge computing device has a certain computing power and storage space, and can filter and compress the original data, retaining only the characteristic data related to the meter reading. By performing anomaly detection analysis on the edge device, the system can mark abnormal data points that exceed the preset threshold range and perform preliminary analysis of the data, such as trend prediction and status assessment. Important data analysis results are uploaded to the cloud via the wireless network for more in-depth analysis and long-term storage. This process not only reduces the network load and improves the real-time response capability of the system, but also ensures the security and integrity of the data, facilitating subsequent query and analysis.

[0082] From the above, it can be seen that the key points of this application include:

[0083] 1. Autonomous navigation inspection technology.

[0084] Technical means:

[0085] Multi-sensor fusion navigation system: Integrates multiple types of sensors such as LiDAR, ultrasonic sensors, cameras, etc. to achieve all-round environmental perception.

[0086] SLAM algorithm optimization: Using an improved SLAM algorithm, the environment map is constructed in real time and the robot position is accurately located to ensure the accuracy and stability of autonomous navigation.

[0087] Path planning and obstacle avoidance algorithms: Combining classic path planning algorithms such as A* and Dijkstra with advanced obstacle avoidance strategies to ensure that the robot moves efficiently and avoids obstacles in complex environments.

[0088] Key points of the invention:

[0089] It achieves high-precision autonomous navigation capabilities and can move flexibly in the narrow and complex environment of the distribution room.

[0090] It provides a reliable obstacle avoidance function to ensure the safe operation of the robot during the inspection process.

[0091] 2. High-precision image acquisition and processing technology.

[0092] Technical means:

[0093] Multi-DOF robotic arm and high-definition camera: Equipped with a high-resolution camera and a multi-DOF robotic arm, it can flexibly adjust the shooting angle and height as needed to ensure that each instrument can be clearly photographed.

[0094] Image preprocessing module: including denoising, contrast enhancement, cropping and other functions to improve image quality and provide reliable data for subsequent recognition.

[0095] Deep learning model: Use convolutional neural network (CNN) for feature extraction and combine it with optical character recognition (OCR) technology to accurately identify numbers and symbols on the instrument.

[0096] Key points of the invention:

[0097] Multi-angle and high-precision image acquisition is achieved to ensure that all key areas are effectively covered.

[0098] Advanced image recognition algorithms are applied to significantly improve the accuracy and reliability of readings.

[0099] 3. Advanced image recognition algorithm technology.

[0100] Technical means:

[0101] Intelligent Character Recognition (OCR): OCR technology based on deep learning can automatically recognize and convert characters in images into readable text.

[0102] Contextual analysis and correction: Combine historical data and contextual information to further analyze and correct the recognition results to improve data accuracy.

[0103] Anomaly detection and alarm: Real-time monitoring of recognition results, and immediate alarm when abnormal situations are found, reminding relevant personnel to handle them.

[0104] Key points of the invention:

[0105] Provides high-precision character recognition capabilities, reducing the need for human intervention.

[0106] The intelligent level of data processing has been enhanced, and automated anomaly detection and early warning have been achieved.

[0107] 4. Data backhaul edge computing technology.

[0108] Technical means:

[0109] Edge computing device: A small computing device deployed locally with certain computing power and storage space for preliminary data processing and analysis.

[0110] Data filtering and compression: Filter and compress raw data on edge devices to upload only useful information and reduce transmission volume.

[0111] Real-time processing and cloud collaboration: Use edge computing capabilities to perform preliminary data analysis, and upload important results to the cloud via wireless networks for more in-depth analysis and long-term storage.

[0112] Key points of the invention:

[0113] Low-latency data processing is achieved, improving the system's real-time response capabilities.

[0114] It saves bandwidth resources, reduces network load, and ensures data security and integrity.

[0115] like Figure 2 and Figure 3 A flowchart provided for an optional embodiment of the present application, wherein: Figure 2 Specifically, it is the specific process from self-checking to image preprocessing. Figure 3 It is a process of further processing of image preprocessing; the invention creates a multi-state photography edge analysis method to achieve the following advantages and positive effects:

[0116] 1. Improve security.

[0117] Existing problem: Manual inspection requires workers to directly enter high-voltage dangerous environments, which poses a risk of electric shock and other accidents.

[0118] Technical means: The robot is equipped with multiple sensors (such as lidar, ultrasonic sensors, etc.) and advanced SLAM algorithms, which can move autonomously and avoid obstacles in complex environments. Through wireless communication technology, operators can monitor and control the robot's actions from a safe distance.

[0119] Advantages and positive effects: It avoids the risk of direct contact between personnel and high-voltage equipment, reducing the possibility of personal injury; the robot can automatically avoid obstacles and ensure safe movement in narrow spaces, further reducing the possibility of accidents.

[0120] 2. Improve work efficiency.

[0121] Existing problem: Manual inspection is time-consuming and laborious, especially in large distribution rooms, it is very time-consuming to check each instrument one by one and record the data.

[0122] Technical means: The robot autonomously navigates according to the preset path, quickly reaches each inspection point, and uses a multi-degree-of-freedom robotic arm and high-definition camera to flexibly adjust the shooting angle and height to complete the shooting task of multiple instruments at one time.

[0123] Advantages and positive effects: Greatly shortens inspection time and improves inspection efficiency; due to faster inspection speed, inspection frequency can be increased to detect potential problems in a timely manner.

[0124] 3. Enhance data accuracy.

[0125] Existing problem: Manual reading of instrument data is prone to errors, especially in poor lighting or unclear instrument displays.

[0126] Technical means: The image preprocessing module includes functions such as denoising, contrast enhancement, and cropping to improve image quality; the deep learning model is combined with optical character recognition (OCR) technology to accurately identify numbers and symbols on the instrument; context analysis and correction combines historical data and context information to further analyze and correct the recognition results.

[0127] Advantages and positive effects: Image quality and recognition accuracy are ensured through image preprocessing and deep learning models; the automated recognition process reduces errors caused by human factors and improves data reliability.

[0128] 4. Optimize data management.

[0129] Existing problems: Existing systems usually lack efficient data management and storage mechanisms, which makes it difficult to preserve and trace data for a long time, affecting subsequent troubleshooting and maintenance work.

[0130] Technical means: Small computing devices deployed locally have certain computing power and storage space for preliminary data processing and analysis; the original data is screened and compressed on the edge device, and only useful information is uploaded to reduce the transmission volume; important data is uploaded to the cloud via wireless network for more in-depth analysis and long-term storage.

[0131] Advantages and positive effects: Local edge computing devices enable fast data processing, reduce data transmission time, and improve the system's real-time response capabilities; selectively upload useful data, reduce network load, and ensure data security and integrity; efficient data management and storage mechanisms facilitate subsequent query and analysis, improving the overall management level.

[0132] Please refer to Figure 4 , which shows a block diagram of a multi-state camera edge analysis system provided by an embodiment of the present application. Figure 4 As shown, the system may include:

[0133] A positioning detection module is used to construct an environmental map in real time using an improved SLAM algorithm and locate the position of the detection robot, and to perform all-round environmental perception of the environment to be detected through multiple types of sensors arranged on the detection robot; wherein the multiple types of sensors include at least a laser radar, an ultrasonic sensor and a camera;

[0134] The acquisition and processing module is used to acquire and preprocess images after the detection robot reaches the designated inspection point; image acquisition includes adjusting the camera shooting angle and height through the preset parameters of the robot arm, and preprocessing includes denoising, contrast enhancement, and cropping the acquired images;

[0135] The feature extraction module is used to extract features from the preprocessed image using a deep learning model and recognize numbers and symbols on instruments in the image in combination with OCR technology;

[0136] The analysis and processing module is used to process and perform preliminary data analysis on the identified data on the edge computing device, and upload the important data analyzed by the data analysis results to the cloud via the wireless network.

[0137] For the specific definition of the multi-state photography edge analysis system, please refer to the definition of the multi-state photography edge analysis method above, which will not be repeated here. Each module in the above-mentioned multi-state photography edge analysis system can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0138] In one embodiment, an electronic device is provided. The electronic device may be a computer, and its internal structure diagram may be as follows: Figure 5 As shown. The electronic device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the device is used to provide computing and control capabilities. The memory of the device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used for multi-state photography edge analysis data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a multi-state photography edge analysis method is implemented.

[0139] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0140] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which involves all or part of the processes in the above-mentioned embodiment method.

[0141] In one embodiment, a computer program product is also provided, including a computer program / instruction, which involves all or part of the process in the above embodiment method.

[0142] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in M ​​forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (SyMchliMk) DRAM (SLDRAM), memory bus (RaMbus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0143] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0144] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the patent application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent application shall be subject to the attached claims.

Claims

1. A multi-stance photography edge analysis method, characterized in that: The method comprises: The improved SLAM algorithm is used to construct an environmental map in real time and locate the position of the detection robot, and the multi-type sensors set on the detection robot are used to perform all-round environmental perception of the environment to be detected; wherein the multi-type sensors include at least a laser radar, an ultrasonic sensor and a camera; When the inspection robot arrives at the designated inspection point, it performs image acquisition and preprocessing. Image acquisition includes adjusting the camera shooting angle and height through the preset parameters of the robot arm, and preprocessing includes denoising, contrast enhancement, and cropping the acquired images. Use deep learning models to extract features from preprocessed images and combine them with OCR technology to identify numbers and symbols on instruments in the images; The identified data is processed and preliminarily analyzed on the edge computing device, and the important data analyzed are uploaded to the cloud via the wireless network.

2. The multi-state photographing edge analysis method according to claim 1, characterized in that: The step of using the improved SLAM algorithm to construct an environment map in real time and locate the position of the detection robot also includes: Combined with A* or Dijkstra path planning algorithms, the optimal path from the current position to the next inspection point is planned for the inspection robot based on the real-time constructed environment map, and the path is dynamically adjusted when encountering obstacles.

3. The multi-state photographing edge analysis method according to claim 1, characterized in that: The steps of image acquisition and preprocessing also include: The multi-degree-of-freedom robotic arm dynamically adjusts the camera's shooting angle and height based on the instrument's position and posture detected in real time to ensure that the instrument is fully displayed in the center area of ​​the captured image.

4. The multi-stance photography edge analysis method according to claim 1, characterized in that: The step of extracting features from the preprocessed image using a deep learning model also includes: A convolutional neural network is used to perform multi-layer feature extraction on the image to extract the key feature areas in the image, including the outline of the dashboard, the position of the pointer, and the shape features of the digital characters.

5. The multi-stance photography edge analysis method according to claim 1, characterized in that: The step of using OCR technology to recognize numbers and symbols on instruments in the image also includes: Combined with the context analysis algorithm, the OCR recognition results are corrected. By comparing the rationality of historical data and current recognition results, recognition errors caused by image quality or character ambiguity can be corrected.

6. The multi-state photographing edge analysis method according to claim 1, characterized in that: The step of performing data processing and preliminary data analysis on the edge computing device after identification also includes: The raw data is filtered and compressed on the edge computing device, retaining only the feature data related to the meter readings, and the data is analyzed for anomaly detection to mark abnormal data points that exceed the preset threshold range.

7. A multi-state photographing edge analysis system, characterized in that: The system comprises: A positioning detection module is used to construct an environmental map in real time using an improved SLAM algorithm and locate the position of the detection robot, and to perform all-round environmental perception of the environment to be detected through multiple types of sensors arranged on the detection robot; wherein the multiple types of sensors include at least a laser radar, an ultrasonic sensor and a camera; The acquisition and processing module is used to acquire and preprocess images after the detection robot reaches the designated inspection point; image acquisition includes adjusting the camera shooting angle and height through the preset parameters of the robot arm, and preprocessing includes denoising, contrast enhancement, and cropping the acquired images; The feature extraction module is used to extract features from the preprocessed image using a deep learning model and recognize numbers and symbols on instruments in the image in combination with OCR technology; The analysis and processing module is used to process and perform preliminary data analysis on the identified data on the edge computing device, and upload the important data analyzed by the data analysis results to the cloud via the wireless network.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images