An auxiliary positioning system based on computer vision recognition and image processing

Abstract

The present invention relates to the technical field of image processing, and discloses an auxiliary positioning system based on computer vision recognition and image processing. The auxiliary positioning system based on computer vision recognition and image processing achieves dual improvement of positioning accuracy and stability through multi-source data acquisition, intelligent data processing, precise calibration and dynamic adjustment. Through comparative analysis of calibration coefficient BSX and multi-level thresholds, it can adapt to complex environmental changes more flexibly. The combination of such adaptability and precise calibration makes the performance of the system more robust and reliable in outdoor and dynamic scenes, thereby significantly improving the accuracy of positioning and recognition and the overall efficiency of the system. A modular calculation method is used to process and calculate different types of data separately, which can avoid data confusion and ensure the accuracy of processing various types of data.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an auxiliary positioning system based on computer vision recognition and image processing. Background Art

[0002] In recent years, computer vision has been widely used in many application scenarios, including autonomous driving, intelligent monitoring, robot navigation, etc. Among them, image processing technology based on computer vision has become a core technical means to solve problems such as positioning, recognition and environmental perception.

[0003] The existing auxiliary positioning systems still have some obvious shortcomings when facing complex environments and changing dynamic scenes. When processing multi-source data, the existing systems often lack effective fusion and calibration mechanisms, which affects the positioning accuracy and system response efficiency. Traditional positioning systems often rely on a single data source for calculations and are easily disturbed by changes in the external environment, resulting in inaccurate positioning or even failure. In addition, when dealing with complex or changing scenes, the existing systems are difficult to adjust and calibrate in real time and lack the ability to respond flexibly; therefore, they do not meet existing needs. In this regard, we propose an auxiliary positioning system based on computer vision recognition and image processing. Summary of the invention

[0004] The present invention provides an auxiliary positioning system based on computer vision recognition and image processing, which has the beneficial effect of greatly improving the accuracy and adaptability of the system in complex scenes through the integrated processing and real-time calibration of multi-source data, so as to overcome the disadvantages of low positioning accuracy and poor environmental adaptability in the prior art, and solve the problem mentioned in the above background technology that the existing system often lacks an effective fusion and calibration mechanism when processing multi-source data, resulting in the positioning accuracy and system response efficiency being affected. Traditional positioning systems often rely on a single data source for calculations, and are easily disturbed by changes in the external environment, resulting in inaccurate or even failure of positioning, and the existing system is difficult to adjust and calibrate in real time when dealing with complex or changing scenes, and lacks the ability to respond flexibly.

[0005] The present invention provides the following technical solutions: an auxiliary positioning system based on computer vision recognition and image processing, the system comprising a data acquisition unit, a data processing unit, a data calculation unit, a data analysis unit, an execution unit and a communication unit;

[0006] The data acquisition unit is used for acquiring multi-source data of the image;

[0007] The data processing unit is used to preprocess the multi-source data collected by the data collection unit, and reorganize the preprocessed multi-source data to generate a first data set, a second data set, and a third data set;

[0008] The data calculation unit is used to perform integrated calculation on the first data set, the second data set and the third data set, so as to generate a calibration coefficient BSX;

[0009] The data analysis unit is used to compare the calibration coefficient BSX with a preset first threshold value Y to generate a first comparison result, and judge whether the current image needs to be calibrated according to the first comparison result. If the judgment result is that calibration is needed, the calibration coefficient BSX is integrated and calculated with the preset first threshold value Y to generate a magnitude coefficient LJZ, and the magnitude coefficient LJZ is compared with a preset second threshold value R to generate a second comparison result;

[0010] The execution unit is used to execute an adaptive calibration measure according to the second comparison result;

[0011] The communication unit is used to upload the device's operating status, environmental parameters and other information to the terminal, and receive instructions and parameter updates from the terminal.

[0012] Preferably, the data acquisition unit includes a first acquisition module, a second acquisition module and a third acquisition module;

[0013] The first acquisition module is used to acquire camera image data and form a camera image data set including edge feature coordinates, pixel depth values ​​and image gradients;

[0014] The second acquisition module is used to acquire spatial positioning data and form a spatial positioning data set including a camera rotation matrix, a camera displacement vector and a camera motion vector;

[0015] The third acquisition module is used to collect environmental data and form an environmental data set including light intensity values, obstacle distances, and temperature values.

[0016] Preferably, the data processing unit includes a preprocessing module and an integration module;

[0017] The preprocessing module is used to preprocess the camera image data set, the spatial positioning data set and the environmental data set, and to dimensionlessly transform the preprocessed multi-source data;

[0018] The integration module divides the image into several regions, denoted as i1, i2, i3, ..., in, and organizes the dimensionless multi-source data according to the region numbers to form the first data set, the second data set, and the third data set.

[0019] Preferably, the first data set includes edge feature coordinates, pixel depth values, and image gradients;

[0020] The edge feature coordinates are denoted as TZB according to the region number. i1 , TZBi2 , TZB i3 ,...,TZB in ;

[0021] The pixel depth values ​​are recorded as XSD according to the region number. i1 , XSD i2 , XSD i3 ,..., XSD in ;

[0022] Image gradients are denoted as TDZ according to the region number i1 , TDZ i2 , TDZ i3 ,...,TDZ in ;

[0023] The second data set includes a camera rotation matrix, a camera displacement vector, and a camera motion vector;

[0024] The camera rotation matrix coordinates are recorded as XXB according to the region number. i1 , XXB i2 , XXB i3 ,...,XXB in ;

[0025] The camera displacement vector is recorded as XWY according to the region number. i1 、XWY i2 、XWY i3 、...、XWY in ;

[0026] The camera motion vector is recorded as XJS according to the region number. i1 、XJS i2 、XJS i3 ,...,XJS in ;

[0027] The third data set includes light intensity values, obstacle distances, and temperature values;

[0028] The light intensity values ​​are recorded as GQD according to the region number. i1 , GQD i2 , GQD i3 ,...,GQD in ;

[0029] Obstacle distances are recorded as ZAW according to the area number. i1 , ZAW i2 , ZAW i3 、...、ZAW in ;

[0030] The temperature values ​​are recorded as WDZ according to the zone number.i1 , WDZ i2 , WDZ i3 , ..., WDZ in .

[0031] Preferably, the data calculation unit includes a first calculation module and a second calculation module;

[0032] The first calculation module is used to integrate and calculate the first data set, the second data set and the third data set into a first intervention value S1, a second intervention value S2 and a correction value K respectively;

[0033] The second calculation module is used for integrating the first intervention value, the second intervention value and the correction value to calculate the calibration coefficient BSX.

[0034] Preferably, the first intervention value S1, the second intervention value S2 and the correction value K are calculated and obtained by the following formulas respectively:

[0035] ;

[0036] ;

[0037] ;

[0038] Where: TZB in , XSD in , TDZ in , XXB in 、XWY in 、XJS in , GQD in , ZAW in and WDZ in They are the edge feature coordinates of the area numbered in, the pixel depth value, the image gradient, the light intensity value, the obstacle distance, the temperature value, the camera rotation matrix, the camera displacement vector and the camera motion vector. ln is the logarithmic function, e is the base function, a1 and a1 are the weight values, and the values ​​of a1 and a2 are adjusted by the user.

[0039] Preferably, the calibration coefficient BSX is calculated by the following formula:

[0040] ;

[0041] Wherein: S1 is the first intervention value, S2 is the second intervention value, K is the correction value, b1 and b2 are weight values, and the values ​​of b1 and b2 are adjusted and set by the user.

[0042] Preferably, the data analysis unit includes a first analysis module and a second analysis module;

[0043] The first analysis module is used to generate a first comparison result, and the second analysis module is used to generate a second comparison result.

[0044] Preferably, the first comparison result is as follows:

[0045] when When , it means that auxiliary positioning is not needed at present;

[0046] when , it means that auxiliary positioning is currently needed.

[0047] Preferably, the second comparison result is as follows:

[0048] when When , it means that it is currently in the first level deviation state;

[0049] when When , it means that it is currently in the second-level deviation state;

[0050] when , it indicates that the current state is in the third level deviation state.

[0051] The present invention has the following beneficial effects:

[0052] 1. This auxiliary positioning system based on computer vision recognition and image processing has achieved a dual improvement in positioning accuracy and stability through multi-source data collection, intelligent data processing, precise calibration and dynamic adjustment. Through comparative analysis of the calibration coefficient BSX and multi-level thresholds, it can more flexibly adapt to complex environmental changes. This combination of adaptability and precise calibration makes the system more robust and reliable in outdoor and dynamic scenes, thereby significantly improving the accuracy of positioning and recognition and the overall efficiency of the system.

[0053] 2. The auxiliary positioning system based on computer vision recognition and image processing adopts a modular calculation method to process and calculate different types of data separately, which can avoid data confusion and ensure the accuracy of various data processing. At the same time, through the first intervention value S1, the second intervention value S2 and the correction value K, this design can flexibly respond to changes in various environments and conditions, ensuring that the system maintains high-precision calculation results in different situations. It has higher flexibility and adaptability in the data integration stage, especially in the face of complex or changing environments. It can balance the influence of multiple data sources through the calculation of different intervention values ​​and correction values ​​to ensure stable and accurate output of the system. By summarizing the first intervention value S1, the second intervention value S2 and the correction value K, a global calibration coefficient BSX is generated, and accurate calibration after multi-source data fusion is achieved. Compared with the traditional single data processing method, the system can comprehensively evaluate the influence of different data sources on the final calibration, thereby ensuring more accurate results. The calibration coefficient BSX calculation brought by multi-source data fusion not only improves the calibration accuracy, but also enhances the reliability of the system in dealing with complex scenes. This modular calculation method can greatly reduce the probability of errors in the system in complex or dynamic environments and improve the overall operation efficiency and accuracy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The figure is a flow chart of the system of the present invention.

[0055] In the figure: 1. Data acquisition unit; 11. First acquisition module; 12. Second acquisition module; 13. Third acquisition module; 2. Data processing unit; 21. Preprocessing module; 22. Integration module; 3. Data calculation unit; 31. First calculation module; 32. Second calculation module; 4. Data analysis unit; 41. First analysis module; 42. Second analysis module; 5. Execution unit; 6. Communication unit. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] Example 1

[0058] See also Figure 1 , an auxiliary positioning system based on computer vision recognition and image processing, the system includes a data acquisition unit 1, a data processing unit 2, a data calculation unit 3, a data analysis unit 4, an execution unit 5 and a communication unit 6;

[0059] The data acquisition unit 1 is used for performing multi-source data acquisition on the image;

[0060] The data processing unit 2 is used to preprocess the multi-source data collected by the data collection unit 1, and reorganize the preprocessed multi-source data to generate a first data set, a second data set, and a third data set;

[0061] The data calculation unit 3 is used to perform integrated calculation on the first data set, the second data set and the third data set, so as to generate a calibration coefficient BSX;

[0062] The data analysis unit 4 is used to compare the calibration coefficient BSX with the preset first threshold value Y to generate a first comparison result, and judge whether the current image needs to be calibrated according to the first comparison result. If the judgment result is that calibration is needed, the calibration coefficient BSX is integrated and calculated with the preset first threshold value Y to generate a magnitude coefficient LJZ, and the magnitude coefficient LJZ is compared with the preset second threshold value R to generate a second comparison result;

[0063] The execution unit 5 is used to execute an adaptive calibration measure according to the second comparison result;

[0064] The communication unit 6 is used to upload information such as the operating status and environmental parameters of the device to the terminal, and to receive instructions and parameter updates from the terminal.

[0065] In this embodiment: Data acquisition unit 1 is responsible for collecting image data of the environment through a variety of sensors including cameras, laser radars, and infrared sensors. By acquiring multi-source data, the system can capture the characteristics of the target object from different angles, increasing the richness and accuracy of the data. The task of this step is to provide a multi-dimensional information basis for subsequent positioning and identification, ensuring that the system can collect sufficiently accurate data in various environments. This not only improves the system's adaptability to complex environments, but also lays a solid foundation for subsequent data processing.

[0066] The data processing unit 2 performs preprocessing on the collected multi-source data, such as noise removal, illumination balancing, edge enhancement, etc., to ensure the quality of the data and make it suitable for further analysis and calculation. At the same time, the reorganized data is divided into a first data set, a second data set, and a third data set. Such classification helps to adopt special algorithms for processing different types of data. The task of this step is to structure the collected complex data to ensure that subsequent calculations can be carried out under orderly and efficient conditions.

[0067] The data calculation unit 3 generates the calibration coefficient BSX by integrating information from different data sets. This coefficient is used to describe the calibration status of the camera or other equipment in three-dimensional space, thereby providing a guarantee for the positioning accuracy of the system. The task of this step is to ensure the calibration accuracy of the equipment based on the integration of multi-source data and avoid positioning errors caused by inconsistent data or equipment deviation. This provides high-precision data support for subsequent analysis and decision-making, and improves the stability and reliability of the system.

[0068] The data analysis unit 4 compares the generated calibration coefficient BSX with the preset first threshold value Y, and determines whether the current image needs to be calibrated based on the comparison result. When it is determined that calibration is required, the system further calculates the magnitude coefficient LJZ and compares it with the second threshold value R to obtain a second comparison result. This analysis process helps to ensure that the system adjusts its own state in real time during operation, automatically detects and corrects potential errors, and thus maintains stable performance. The task of this step is to dynamically detect the operating state of the equipment, automatically adjust to ensure accuracy, and improve the system's adaptive ability.

[0069] The execution unit 5 performs adaptive calibration measures based on the second comparison result output by the data analysis unit. When it is detected that the device or image needs to be calibrated, the system will automatically take measures, such as adjusting the camera's posture or recalibrating the device position, to ensure the accuracy of positioning. The task of this step is to perform calibration operations based on the analysis results, thereby effectively reducing the errors of positioning and recognition and improving the overall reliability and adaptability of the system.

[0070] The communication unit 6 is responsible for uploading the system's operating status, environmental parameters and other information to the terminal, and receiving instructions and parameter updates issued by the terminal. This step ensures real-time interaction between the system and the external environment, allowing users or monitoring systems to understand the device status at any time and make remote adjustments or issue instructions as needed. The completion of this task enables the system to have the ability of remote control and monitoring, thereby improving the flexibility and response speed of the system.

[0071] Compared with traditional positioning systems, this system has achieved a dual improvement in positioning accuracy and stability through multi-source data collection, intelligent data processing, precise calibration and dynamic adjustment. Compared with current technical means, this system performs well in data preprocessing and real-time calibration. Through comparative analysis of the calibration coefficient BSX and multi-level thresholds, it can adapt to complex environmental changes more flexibly. This combination of adaptability and precise calibration makes the system more robust and reliable in outdoor and dynamic scenes, thereby significantly improving the accuracy of positioning and recognition and the overall efficiency of the system.

[0072] Example 2

[0073] See also Figure 1 , the data acquisition unit 1 includes a first acquisition module 11, a second acquisition module 12 and a third acquisition module 13;

[0074] The first acquisition module 11 is used to acquire camera image data and form a camera image data set including edge feature coordinates, pixel depth values ​​and image gradients;

[0075] The second acquisition module 12 is used to acquire spatial positioning data and form a spatial positioning data set including a camera rotation matrix, a camera displacement vector and a camera motion vector;

[0076] The third acquisition module 13 is used to collect environmental data and form an environmental data set including light intensity values, obstacle distances and temperature values.

[0077] In this embodiment: by collecting camera image data, including edge feature coordinates, pixel depth values, and image gradients, the system can accurately capture the multi-dimensional features of the environment, providing rich data support for subsequent image processing and object recognition. Edge feature coordinates and image gradient data can effectively enhance the system's ability to recognize object contours and textures, especially in complex backgrounds, improving recognition accuracy. At the same time, pixel depth values ​​enable the system to convert two-dimensional image data into three-dimensional spatial information, thereby enhancing the perception of spatial distance and depth, ensuring that the system has higher accuracy and stability during navigation and positioning.

[0078] By collecting the camera rotation matrix, displacement vector and motion vector, the system can accurately track the motion state of the camera or device. Whether it is translation or rotation, the system can accurately grasp the posture change of the device. Through the camera motion vector data, the system can dynamically compensate for the image deviation caused by camera movement, thereby ensuring the continuity and stability of positioning. The collection of this data enables the system to calculate the relative position between the device and the target object, helping the system to achieve high-precision spatial positioning and target tracking.

[0079] By collecting light intensity values, the system can automatically adjust the processing algorithm under different lighting conditions to ensure image quality and recognition stability, and can work normally in both strong and weak light environments. The collection of obstacle distance data enables the system to have real-time environmental perception and obstacle detection capabilities, and can actively avoid potential collision risks, which is especially suitable for complex outdoor scenes. The collection of temperature values ​​allows the system to dynamically perceive the ambient temperature and compensate for the impact of temperature on device performance, such as sensor drift, to ensure stable operation of the device under different temperature conditions.

[0080] Example 3

[0081] See also Figure 1, the data processing unit 2 includes a preprocessing module 21 and an integration module 22;

[0082] The preprocessing module 21 is used to preprocess the camera image data set, the spatial positioning data set and the environment data set, and to dimensionlessly convert the preprocessed multi-source data;

[0083] The integration module 22 divides the image into a number of regions, respectively denoted as i1, i2, i3, . . . , in, and sorts the dimensionless multi-source data according to the region numbers to form a first data set, a second data set, and a third data set.

[0084] The first data set includes edge feature coordinates, pixel depth values, and image gradients;

[0085] The edge feature coordinates are denoted as TZB according to the region number. i1 , TZB i2 , TZB i3 ,...,TZB in ;

[0086] The pixel depth values ​​are recorded as XSD according to the region number. i1 , XSD i2 , XSD i3 ,..., XSD in ;

[0087] Image gradients are denoted as TDZ according to the region number i1 , TDZ i2 , TDZ i3 ,...,TDZ in ;

[0088] The second data set includes a camera rotation matrix, a camera displacement vector, and a camera motion vector;

[0089] The camera rotation matrix coordinates are recorded as XXB according to the region number. i1 , XXB i2 , XXB i3 ,...,XXB in ;

[0090] The camera displacement vector is recorded as XWY according to the region number. i1 、XWY i2 、XWY i3 、...、XWY in ;

[0091] The camera motion vector is recorded as XJS according to the region number. i1 、XJS i2 、XJS i3 ,...,XJSin ;

[0092] The third data set includes light intensity values, obstacle distances, and temperature values;

[0093] The light intensity values ​​are recorded as GQD according to the region number. i1 , GQD i2 , GQD i3 ,...,GQD in ;

[0094] Obstacle distances are recorded as ZAW according to the area number. i1 , ZAW i2 , ZAW i3 、...、ZAW in ;

[0095] The temperature values ​​are recorded as WDZ according to the zone number. i1 , WDZ i2 , WDZ i3 , ..., WDZ in。

[0096] In this embodiment: the preprocessing module 21 performs preliminary processing on the camera image data set, the spatial positioning data set and the environmental data set, such as noise filtering, abnormal data correction, etc., to ensure the integrity and quality of the data and provide a reliable basis for subsequent processing. By dimensionlessly processing the preprocessed multi-source data, the influence between different data types and dimensions is eliminated, so that the system can more flexibly handle the integration and analysis of various different data sources, improve data interoperability and computing efficiency. After dimensionlessly processing, all data are uniformly analyzed under the same dimension, reducing the deviation caused by the difference in data magnitude, facilitating subsequent accurate calculation and analysis, and improving the consistency of data processing.

[0097] The integration module 22 divides the image into several regions i1, i2, i3, ..., in, and each region is independently numbered and processed. This partition processing method can improve the focus on local features, help the system improve the accuracy of recognition and positioning in scenes with complex details, and organize the dimensionless multi-source data according to the region number. The system can achieve refined multi-source data integration, thereby effectively organizing and managing information from different data sources, ensuring that multi-dimensional data can be accurately integrated, and through the processing of the integration module, forming the first data set, the second data set, and the third data set, providing a clear and orderly structure for the subsequent calculation unit. Data input improves the overall data processing efficiency and system response speed.

[0098] Through preprocessing and dimensionless operations, the consistency and quality of different types of data are guaranteed, the errors in system processing are reduced, and the foundation is laid for accurate positioning and identification. Regional segmentation makes the system's image processing in complex scenes more refined, which helps to improve the system's recognition and positioning accuracy of target objects, especially suitable for environments with complex details. Through reasonable regional division and multi-source data collation, the system can more effectively organize and manage large amounts of input data, improve the efficiency and accuracy of data processing, and ultimately improve the performance and response speed of the entire system.

[0099] By collecting edge feature coordinates, pixel depth values, and image gradients in a regional numbering manner, the system can more accurately locate the edge features and depth information of objects, thereby improving the accuracy of object recognition, especially in scenes with complex environments, enhancing the ability to recognize object shapes, contours, and textures. Combined with image gradients and pixel depth values, the system can achieve the conversion of two-dimensional images to three-dimensional space, and further perceive the spatial relationship between objects and the environment. This multi-dimensional perception capability significantly improves the system's understanding of the position, size, and relative distance of target objects, providing more accurate data support for the system's subsequent positioning and calibration. The precise extraction of edge feature coordinates and image gradients can help the system maintain high recognition performance in complex backgrounds, avoid recognition deviations caused by background interference, and thus maintain high recognition efficiency in outdoor or irregular environments.

[0100] By processing the camera rotation matrix, camera displacement vector and camera motion vector into regional numbers, the system can better track the camera's motion state and position changes, thereby improving the spatial positioning accuracy of the entire scene. Especially in dynamic environments, this spatial positioning capability is crucial for auxiliary positioning systems. The introduction of camera motion vectors enables the system to track dynamic objects. By combining motion data, the system can adjust the recognition and positioning of objects in real time, enhancing adaptability and robustness in dynamic scenes. By integrating the camera's rotation matrix and displacement vector, the system can accurately calculate the relative position in the entire scene, ensuring that the spatial relationship between the object and the environment can be accurately described, and through multi-source data fusion, significantly improving the accuracy and stability of the entire positioning system.

[0101] The collection of light intensity values ​​can help the system dynamically adjust its performance under different lighting conditions, ensuring that the system can maintain high recognition and positioning accuracy in scenes with drastic lighting changes, thereby improving the environmental adaptability of the system. Through real-time collection of obstacle distances, the system can effectively identify obstacles in the surrounding environment and make corresponding avoidance decisions, thereby increasing the system's security and intelligent processing capabilities, and is especially suitable for complex navigation and obstacle avoidance scenarios. The collection of temperature values ​​enables the system to monitor changes in ambient temperature, ensuring that the system maintains normal operation in scenes with large temperature fluctuations, preventing system performance fluctuations or failures caused by abnormal temperature, and further improving the stability of the system.

[0102] Through the orderly partitioned collection and organization of the first, second and third data sets, the system can efficiently integrate multi-source data, thereby achieving significant improvements in accuracy and performance in recognition, positioning, environmental perception and other aspects. The mutual cooperation between various data sets enables the system to not only achieve high-precision positioning and recognition in static scenes, but also have the ability to respond quickly and adaptively adjust in dynamic and complex environments. By monitoring environmental data such as light intensity, temperature and obstacle distance, the system can dynamically adjust the working state according to changes in environmental conditions, ensuring the stable operation of the system under different environmental conditions, and effectively improving the robustness and reliability of the overall system.

[0103] Example 4

[0104] See also Figure 1 , the data calculation unit 3 includes a first calculation module 31 and a second calculation module 32;

[0105] The first calculation module 31 is used to integrate and calculate the first data set, the second data set and the third data set into a first intervention value S1, a second intervention value S2 and a correction value K respectively;

[0106] The second calculation module 32 is used for integrating the first intervention value, the second intervention value and the correction value to calculate the calibration coefficient BSX.

[0107] The first intervention value S1, the second intervention value S2 and the correction value K are calculated and obtained by the following formulas respectively;

[0108] ;

[0109] ;

[0110] ;

[0111] Where: TZB in , XSD in , TDZ in , XXB in 、XWYin 、XJS in , GQD in , ZAW in and WDZ in They are the edge feature coordinates of the area numbered in, the pixel depth value, the image gradient, the light intensity value, the obstacle distance, the temperature value, the camera rotation matrix, the camera displacement vector and the camera motion vector. ln is the logarithmic function, e is the base function, a1 and a1 are the weight values, and the values ​​of a1 and a2 are adjusted by the user.

[0112] The calibration coefficient BSX is calculated by the following formula:

[0113] ;

[0114] Wherein: S1 is the first intervention value, S2 is the second intervention value, K is the correction value, b1 and b2 are weight values, and the values ​​of b1 and b2 are adjusted and set by the user.

[0115] In this embodiment: This system uses a modular calculation method to process and calculate different types of data separately, which can avoid data confusion and ensure the accuracy of various data processing. At the same time, through the first intervention value S1, the second intervention value S2 and the correction value K, this design can flexibly respond to changes in various environments and conditions, ensuring that the system maintains high-precision calculation results in different situations, and has higher flexibility and adaptability in the data integration stage. Especially in the face of complex or changing environments, it can balance the influence of multiple data sources through the calculation of different intervention values ​​and correction values ​​to ensure stable and accurate output of the system.

[0116] By summarizing the first intervention value S1, the second intervention value S2 and the correction value K, a global calibration coefficient BSX is generated, achieving accurate calibration after multi-source data fusion. Compared with the traditional single data processing method, the system can comprehensively evaluate the impact of different data sources on the final calibration, thereby ensuring more accurate results. The calibration coefficient BSX calculation brought by multi-source data fusion not only improves the calibration accuracy, but also enhances the reliability of the system when dealing with complex scenarios. This modular calculation method can greatly reduce the probability of errors in the system in complex or dynamic environments, and improve the overall operation efficiency and accuracy of the system.

[0117] Through the first calculation module 31 and the second calculation module 32 in the data calculation unit 3, the system has achieved significant improvements in multi-source data processing. Through the step-by-step calculation and integration of intervention values ​​and correction values, the system can process information from different sensors and data sources more finely, ensure the effective fusion of multi-source data, and finally generate a high-precision calibration coefficient BSX. Compared with traditional image recognition and positioning technology, this system not only improves recognition accuracy and environmental adaptability, but also can cope with variable factors in dynamic and complex environments. These improvements have significantly improved the accuracy, reliability and efficiency of the system, and are particularly suitable for high-demand positioning and calibration scenarios.

[0118] Compared with traditional image recognition methods, the first intervention value S1 formula is based on region numbering and jointly processes image edge features with depth and gradient data, which can significantly improve the image's spatial information resolution capabilities, especially under complex backgrounds or lighting conditions. It can more accurately identify objects. This integration method improves the accuracy of image processing, especially for the precise identification of object contours and spatial positions, enabling the positioning system to achieve more reliable target recognition and tracking in more complex environments.

[0119] The second intervention value S2 integrates the light intensity, obstacle distance and temperature data in the environment. This formula enhances the system's sensitivity to environmental changes. For example, in low-light or high-temperature environments, this formula can help the system better understand the environmental conditions of the current scene, thereby providing a more accurate basis for target recognition. By incorporating environmental factors into the calculation, the system has a stronger ability to adapt to the environment and can maintain positioning accuracy under different environmental conditions. It is particularly suitable for complex outdoor or industrial scenes, reducing the impact of environmental interference on recognition and positioning.

[0120] The correction value K can dynamically adjust the calibration result of the system by using logarithmic and exponential functions and combining the motion parameters of the camera. In particular, when the system moves or rotates, the correction value can reflect the changes in the camera state in real time to ensure the accuracy of the positioning result. The introduction of the correction value enables the system to have real-time adjustment capabilities. When the camera moves or rotates, the system can automatically adjust according to dynamic changes to maintain high-precision positioning and recognition.

[0121] By adjusting the weight value, the formula allows users to customize the influence of different data in the calibration process according to specific scenarios and needs. This flexibility significantly improves the system's ability to adapt to different application scenarios. The adjustable weight value enables the system to achieve personalized optimization for different environments and conditions, ensuring the calibration accuracy of the system in various scenarios and improving the versatility and practicality of the system.

[0122] By integrating and calculating multi-source data such as edge feature coordinates, depth values, image gradients, environmental factors, and camera motion parameters, this system has achieved significant breakthroughs in multi-source data fusion and dynamic calibration. Compared with traditional positioning and calibration methods, this system improves the accuracy and robustness of target recognition and positioning through sophisticated formula design and multi-dimensional data processing. In particular, through flexible and adjustable weight values ​​and correction values, the system can be dynamically adjusted according to actual needs, giving it stronger adaptability in complex environments. Overall, the system has been significantly improved in terms of accuracy, real-time, and environmental adaptability, providing reliable technical support for various high-precision positioning scenarios.

[0123] Example 5

[0124] See also Figure 1 , the data analysis unit 4 includes a first analysis module 41 and a second analysis module 42;

[0125] The first analysis module 41 is used to generate a first comparison result, and the second analysis module 42 is used to generate a second comparison result.

[0126] The first comparison results are as follows:

[0127] when When , it means that auxiliary positioning is not needed at present;

[0128] when , it means that auxiliary positioning is currently needed.

[0129] The second comparison results are as follows:

[0130] when When , it means that it is currently in the first level deviation state;

[0131] when When , it means that it is currently in the second-level deviation state;

[0132] when , it means it is currently in the third level deviation state.

[0133] In this embodiment: the first analysis module 41 is used to generate a first comparison result, and its logical judgment condition is when When , it means that no auxiliary positioning is currently required. When , it means that auxiliary positioning is currently needed. By comparing the calibration coefficient BSX and the first threshold Y, the system can accurately judge the current environmental status and realize the automatic triggering of auxiliary positioning. This automatic judgment mechanism can reduce manual intervention, improve the intelligence level of the system, ensure that the auxiliary positioning function is triggered only when it is actually needed, and avoid unnecessary calculations and resource waste. This mechanism optimizes the resource utilization efficiency of the system, reduces unnecessary system load through intelligent judgment, and thus improves response speed and overall performance. At the same time, it can respond flexibly in a dynamic environment, ensuring that auxiliary positioning is only performed when needed, which increases the intelligence and adaptability of the system.

[0134] The second analysis module 42 is used to generate the second comparison result. The specific logic is as follows: When , it means it is in the first level deviation state; when When , it means that it is currently in the second level deviation state; when , it means that the current state is in the third level of deviation. By judging the deviation state by level, the system can refine and manage the degree of target deviation in a hierarchical manner. This hierarchical judgment mechanism can provide a more detailed reference basis for subsequent positioning corrections, effectively improving the system's ability to cope with complex environments. The classification of deviation states can not only help the system identify and respond to positioning errors more accurately, but also provide different response strategies for different degrees of deviation. Through the refined management of positioning deviations, the system can correct deviations more efficiently and ensure high-precision positioning in the case of different degrees of deviation.

[0135] ; Where: BSX is the calibration coefficient, Y is the first threshold.

[0136] This system further improves the intelligence level and flexibility of the system through the first analysis module 41 and the second analysis module 42 in the data analysis unit 4. The first analysis module 41 can effectively reduce unnecessary resource consumption and ensure that the system provides auxiliary positioning function at the right time through automatic judgment based on the comparison results. The second analysis module 42 provides refined error management through graded judgment of positioning deviations, enhancing the system's ability to cope with complex environments and deviations. Overall, the system shows outstanding improvements and advantages in intelligent judgment, resource optimization and precise positioning management, and provides a reliable, accurate and efficient solution for computer vision-assisted positioning in complex environments.

[0137] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0138] The above are only preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. An auxiliary positioning system based on computer vision recognition and image processing, characterized in that: The system comprises a data acquisition unit (1), a data processing unit (2), a data calculation unit (3), a data analysis unit (4), an execution unit (5) and a communication unit (6); The data acquisition unit (1) is used to acquire multi-source data of the image; The data processing unit (2) is used to pre-process the multi-source data collected by the data collection unit (1), and to reorganize the pre-processed multi-source data, thereby generating a first data set, a second data set, and a third data set; The data calculation unit (3) is used to perform integrated calculation on the first data set, the second data set and the third data set, so as to generate a calibration coefficient BSX; The data analysis unit (4) is used to compare the calibration coefficient BSX with a preset first threshold value Y to generate a first comparison result, and judge whether the current image needs to be calibrated according to the first comparison result. If the judgment result is that calibration is needed, the calibration coefficient BSX and the preset first threshold value Y are integrated and calculated to generate a magnitude coefficient LJZ, and the magnitude coefficient LJZ is compared with a preset second threshold value R to generate a second comparison result; The execution unit (5) is used to execute an adaptive calibration measure according to the second comparison result; The communication unit (6) is used to upload information such as the operating status and environmental parameters of the device to the terminal, and to receive instructions and parameter updates from the terminal; The data acquisition unit (1) comprises a first acquisition module (11), a second acquisition module (12) and a third acquisition module (13); The first acquisition module (11) is used to acquire camera image data and form a camera image data set, including edge feature coordinates, pixel depth values, and image gradients; The second acquisition module (12) is used to acquire spatial positioning data and form a spatial positioning data set, including a camera rotation matrix, a camera displacement vector and a camera motion vector; The third acquisition module (13) is used to collect environmental data and form an environmental data set, including light intensity values, obstacle distances, and temperature values; The data processing unit (2) comprises a preprocessing module (21) and an integration module (22); The preprocessing module (21) is used to preprocess the camera image data set, the spatial positioning data set and the environment data set, and to dimensionlessly convert the preprocessed multi-source data; The integration module (22) divides the image into a plurality of regions, respectively denoted as i1, i2, i3, ..., in, and organizes the dimensionless multi-source data according to the region numbers, thereby forming a first data set, a second data set, and a third data set; The first data set includes edge feature coordinates, pixel depth values, and image gradients; The edge feature coordinates are denoted as TZB according to the region number. i1 , TZB i2 , TZB i3 ,...,TZB in ; The pixel depth values ​​are recorded as XSD according to the area number. i1 , XSD i2 , XSD i3 ,..., XSD in ; Image gradients are denoted as TDZ according to the region number i1 , TDZ i2 , TDZ i3 ,...,TDZ in ; The second data set includes a camera rotation matrix, a camera displacement vector, and a camera motion vector; The camera rotation matrix coordinates are recorded as XXB according to the region number. i1 , XXB i2 , XXB i3 ,...,XXB in ; The camera displacement vector is recorded as XWY according to the region number. i1 、XWY i2 、XWY i3 、...、XWY in ; The camera motion vector is recorded as XJS according to the region number. i1 、XJS i2 、XJS i3 ,...,XJS in ; The third data set includes light intensity values, obstacle distances, and temperature values; The light intensity values ​​are recorded as GQD according to the region number. i1 , GQD i2 , GQD i3 、...、GQD in ; Obstacle distances are recorded as ZAW according to the area number. i1 , ZAW i2 , ZAW i3 、...、ZAW in ; The temperature values ​​are recorded as WDZ according to the zone number. i1 , WDZ i2 , WDZ i3 , ..., WDZ in ; The data calculation unit (3) comprises a first calculation module (31) and a second calculation module (32); The first calculation module (31) is used to integrate and calculate the first data set, the second data set and the third data set into a first intervention value S1, a second intervention value S2 and a correction value K respectively; The second calculation module (32) is used to integrate the first intervention value, the second intervention value and the correction value to calculate the calibration coefficient BSX; The first intervention value S1, the second intervention value S2 and the correction value K are calculated and obtained by the following formulas respectively; Where: TZB in , XSD in , TDZ in , XXB in 、XWY in 、XJS in , GQD in , ZAW in and WDZ in They are the edge feature coordinates, pixel depth value, image gradient, light intensity value, obstacle distance, temperature value, camera rotation matrix, camera displacement vector and camera motion vector of the area numbered in, respectively. ln is a logarithmic function, e is a base function, a1 and a1 are weight values, and the values ​​of a1 and a2 are adjusted by the user; The calibration coefficient BSX is calculated by the following formula: Wherein: S1 is the first intervention value, S2 is the second intervention value, K is the correction value, b1 and b2 are weight values, and the values ​​of b1 and b2 are adjusted and set by the user.

2. The auxiliary positioning system based on computer vision recognition and image processing according to claim 1, characterized in that: The data analysis unit (4) comprises a first analysis module (41) and a second analysis module (42); The first analysis module (41) is used to generate a first comparison result, and the second analysis module (42) is used to generate a second comparison result.

3. The auxiliary positioning system based on computer vision recognition and image processing according to claim 2, characterized in that: The first comparison results are as follows: when When , it means that auxiliary positioning is not needed at present; when , it means that auxiliary positioning is currently needed.

4. The auxiliary positioning system based on computer vision recognition and image processing according to claim 2, characterized in that: The second comparison results are as follows: when When , it means that it is currently in the first level deviation state; when When , it means that it is currently in the second-level deviation state; when , it means it is currently in the third level deviation state.

Images