An instrument data recognition method, device, equipment and storage medium

Abstract

The application discloses an instrument data recognition method, and relates to the technical field of industrial automation. In the method, multi-dimensional image data and structural parameters of an instrument to be detected are collected, the multi-dimensional image data comprises point cloud data and image data, the structural parameters comprise a range of scales, a scale distribution and a display mode; the multi-dimensional image data is subjected to deep layered processing, and instrument layer image data is extracted; the instrument layer image data is subjected to enhancement processing, and final image data is obtained; a three-dimensional model of the instrument to be detected is generated based on the final image data and the structural parameters; key features of the three-dimensional model are extracted and matched with standard features in a preset instrument library, and a matching result is obtained; and according to the matching result, a detection model corresponding to the instrument to be detected is called to perform data recognition on the instrument to be detected. The application improves the accuracy of instrument data recognition.

Description

Technical Field

[0001] This application relates to the field of industrial automation technology, and in particular to a method, apparatus, device and storage medium for instrument data identification. Background Technology

[0002] With the deepening of industrial automation and digital transformation, instrument data identification and acquisition in industrial sites are becoming an important foundation for production process monitoring and intelligent decision-making. In industrial production environments, a large number of instruments such as pressure gauges, flow meters, and thermometers are distributed. These instruments undertake the key task of real-time monitoring of production parameters, and the accurate identification of their data directly affects the safety and controllability of the production process. Therefore, there is an urgent need for an instrument data identification method.

[0003] Existing technologies mainly acquire instrument image data and use image processing methods to identify the image data in order to obtain the instrument's measurement data.

[0004] However, due to the complex environment of industrial sites, the instrument surface is often subject to interference factors such as reflection, stains, and blurring. Traditional single image processing methods are difficult to cope with these complex situations, which can easily lead to large deviations in the recognition results, thus resulting in low accuracy of the instrument data. Summary of the Invention

[0005] This application provides a method, apparatus, device, and storage medium for instrument data identification, which improves the accuracy of instrument data identification.

[0006] A first aspect of this application provides a method for identifying instrument data, comprising: acquiring multidimensional image data and structural parameters of an instrument to be tested, wherein the multidimensional image data includes point cloud data and image data, and the structural parameters include measurement range, scale distribution, and display mode; performing deep layering processing on the multidimensional image data to extract instrument layer image data; performing enhancement processing on the instrument layer image data to obtain final image data; generating a three-dimensional model of the instrument to be tested based on the final image data and the structural parameters; extracting key features of the three-dimensional model and matching them with standard features in a preset instrument library to obtain a matching result; and, based on the matching result, calling the detection model corresponding to the instrument to be tested to perform data identification on the instrument to be tested.

[0007] By adopting the above technical solution, this application obtains multi-dimensional image data by simultaneously acquiring point cloud data and image data. Combined with structural parameters such as range, scale distribution, and display method, it provides rich input information for instrument identification. Deep layering processing is performed on the multi-dimensional image data to extract the instrument layer image data, effectively removing the influence of background and other interference layers. Enhancement processing is applied to the extracted instrument layer image data to solve image quality problems caused by interference factors such as surface reflection, stains, and blurring in industrial environments. Based on the processed high-quality image data and structural parameters, a three-dimensional model of the instrument to be tested is generated, preserving the spatial structural features of the instrument. By extracting key features of the three-dimensional model and matching them with standard features in a preset instrument library, the instrument type is accurately identified. Based on the matching results, the corresponding detection model is called for data identification, achieving accurate identification of instrument data. This solution, through multi-dimensional data acquisition, deep layering extraction, image enhancement processing, three-dimensional model construction, and feature matching, effectively solves the problem of low accuracy in identification using traditional single image processing methods in complex industrial environments, significantly improving the accuracy of instrument data identification.

[0008] Optionally, the step of performing depth layering processing on the multidimensional image data to obtain instrument layer image data specifically includes: constructing a depth mapping matrix based on the multidimensional image data; performing layering processing on the multidimensional image data according to the depth distribution characteristics in the depth mapping matrix to obtain preliminary layering results; performing denoising and fracture region repair processing on each depth layer in the preliminary layering results to obtain optimized layering results; extracting feature sets of each depth layer in the optimized layering results, calculating the similarity between the feature sets of each depth layer and the standard feature set, and selecting the depth layer with the highest similarity as the instrument layer; and converting the instrument layer into a two-dimensional image to obtain instrument layer image data.

[0009] By adopting the above technical solution, this application constructs a depth mapping matrix to perform spatial modeling of the image. The rows and columns of the matrix correspond to the vertical and horizontal coordinates of the image, and the matrix elements record the depth values ​​of the pixels, thereby achieving an accurate description of the spatial structure of the image. The multidimensional image data is processed in layers by using the depth distribution features in the depth mapping matrix. Independent three-dimensional point cloud subsets are generated as depth layers by depth threshold segmentation. Denoising and fracture repair optimization processing is performed on each depth layer to improve the tomography quality. The optimal instrument layer is selected by using feature set similarity calculation and matching. Finally, the three-dimensional instrument layer is converted into two-dimensional image data, thereby achieving effective extraction of the instrument layer in complex industrial environments.

[0010] Optionally, the step of performing layered processing on the multidimensional image data according to the depth distribution characteristics in the depth mapping matrix to obtain a preliminary layering result, wherein the preliminary layering result consists of at least one depth layer, and the depth layer is an independent 3D point cloud subset generated by depth threshold segmentation, specifically including: dividing the data in the multidimensional image data whose depth values ​​are in the same interval into the same layer according to a preset depth threshold and the depth value distribution characteristics in the depth mapping matrix to obtain a preliminary hierarchical structure; calculating the depth difference of adjacent point cloud data in each layer based on the preliminary hierarchical structure, further subdividing each layer based on the depth difference to obtain multiple candidate depth layers; and merging the candidate depth layers based on a preset depth similarity judgment criterion to obtain the preliminary layering result.

[0011] By adopting the above technical solution, this application divides the data with depth values ​​in the same interval in multidimensional image data into the same layer based on the depth value distribution characteristics in the depth mapping matrix and the preset threshold. The initial layering of scene data is achieved by setting multi-level thresholds. The depth difference of adjacent point cloud data in each layer is calculated based on the initial hierarchical structure and inter-layer subdivision is performed to obtain candidate depth layers with similar depth features. The candidate depth layers are merged by using the preset depth similarity judgment criteria, thereby realizing the fine processing of depth layers. The precise division and merging of different depth layers is achieved by combining depth threshold segmentation and depth difference calculation.

[0012] Optionally, the step of extracting feature sets of each depth layer in the optimized layering result, calculating the similarity between the feature sets of each depth layer and the standard feature set, and selecting the depth layer with the highest similarity as the instrument layer specifically includes: extracting feature sets of each depth layer in the optimized layering result, wherein each feature set includes geometric shape features and depth distribution features, and the depth distribution features are distribution features obtained by statistical analysis of depth values; comparing each feature in each feature set with the corresponding features in the standard feature set to determine the initial score corresponding to each feature; determining the similarity score of each feature set based on the initial score and the corresponding preset weight; and selecting the depth layer with the highest similarity score as the instrument layer.

[0013] By adopting the above technical solution, this application constructs a feature set containing geometric shape features and depth distribution features. The geometric shape features reflect the outer contour and internal structure information of the instrument, while the depth distribution features reflect the spatial distribution pattern of the instrument surface through statistical analysis of depth values. The features in each feature set are compared with the standard feature set in multiple dimensions to obtain an initial score. The scores of different features are weighted using preset weights to obtain a similarity score. The optimal instrument layer is selected based on the score results. Feature scoring is achieved through multi-dimensional comparison and weight calculation of features. The depth layer is comprehensively evaluated using a multi-feature fusion method, thus achieving accurate selection of the optimal instrument layer.

[0014] Optionally, the enhancement processing of the instrument layer image data to obtain the final image data specifically includes: performing illumination compensation on the instrument layer image data to obtain first image data; performing edge sharpening processing on the first image data to obtain second image data; and combining the scale distribution to enhance the corresponding areas of the second image data to obtain the final image data.

[0015] By adopting the above technical solution, this application performs multi-stage enhancement processing on the instrument layer image data. First, the uneven illumination on the instrument surface is corrected by illumination compensation technology to obtain the first image data with balanced illumination. The edge sharpening processing of the first image data enhances the edge features of the instrument pointer and scale to obtain the second image data. Combined with the pre-acquired scale distribution information, the scale area is specifically enhanced to obtain the final image data. Through illumination compensation, edge sharpening and region enhancement processing, the instrument layer image is comprehensively optimized, providing high-quality image data for subsequent 3D modeling and data recognition.

[0016] Optionally, generating a three-dimensional model of the instrument to be tested based on the final image data and the structural parameters specifically includes: constructing an initial three-dimensional model based on the point cloud data; extracting a first texture feature from the final image data; constructing constraints according to the measurement range and the display method; registering the initial three-dimensional model with the base model with the highest matching degree in a preset instrument base template library to obtain a registered three-dimensional model; and optimizing the registered three-dimensional model by combining the texture feature and the constraints to obtain a three-dimensional model.

[0017] By adopting the above technical solution, this application constructs an initial three-dimensional model of the instrument based on high-quality point cloud data, extracts texture features of optimized image data as an important basis for model optimization, sets constraints for model optimization according to the instrument's measurement range and display method, achieves model standardization by registering with the basic model in the preset template library, and optimizes and adjusts the registered model in combination with texture features and constraints to achieve accurate reconstruction of the instrument's three-dimensional structure.

[0018] Optionally, the step of extracting key features of the 3D model and matching them with standard features in a preset instrument library to obtain a matching result specifically includes: extracting geometric features, second texture features, and structural features of the 3D model; using the geometric features and the structural features to match based on the corresponding standard geometric features and standard structural features in the preset instrument library to obtain a first matching result; and using the second texture features to match with the corresponding standard texture features in the first matching result to obtain a matching result.

[0019] By adopting the above technical solution, this application performs multi-dimensional feature extraction on the three-dimensional model, and obtains geometric features reflecting the shape of the instrument, texture features reflecting the surface information of the instrument, and structural features reflecting the composition of the instrument. A two-stage matching method is adopted. First, a coarse matching is performed using geometric features and structural features to obtain preliminary results. Then, a precise matching is performed using texture features to obtain the final matching result. Through multi-feature extraction and staged matching, accurate identification of the instrument type is achieved.

[0020] A second aspect of this application provides an instrument data identification system, comprising:

[0021] The multidimensional data acquisition module is used to acquire multidimensional image data and structural parameters of the instrument under test. The multidimensional image data includes point cloud data and image data, and the structural parameters include measurement range, scale distribution and display method.

[0022] The deep layering processing module is used to perform deep layering processing on the multidimensional image data and extract the instrument layer image data;

[0023] An image enhancement processing module is used to enhance the image data of the instrument layer to obtain the final image data;

[0024] A 3D modeling generation module is used to generate a 3D model of the instrument to be tested based on the final image data and the structural parameters.

[0025] The feature matching and analysis module is used to extract key features of the three-dimensional model and match them with standard features in a preset instrument library to obtain matching results;

[0026] The intelligent detection invocation module is used to invoke the detection model corresponding to the instrument under test to perform data identification on the instrument under test based on the matching result.

[0027] By adopting the above technical solution, this application obtains multi-dimensional image data by simultaneously acquiring point cloud data and image data. Combined with structural parameters such as range, scale distribution, and display method, it provides rich input information for instrument identification. Deep layering processing is performed on the multi-dimensional image data to extract the instrument layer image data, effectively removing the influence of background and other interference layers. Enhancement processing is applied to the extracted instrument layer image data to solve image quality problems caused by interference factors such as surface reflection, stains, and blurring in industrial environments. Based on the processed high-quality image data and structural parameters, a three-dimensional model of the instrument to be tested is generated, preserving the spatial structural features of the instrument. By extracting key features of the three-dimensional model and matching them with standard features in a preset instrument library, the instrument type is accurately identified. Based on the matching results, the corresponding detection model is called for data identification, achieving accurate identification of instrument data. This solution, through multi-dimensional data acquisition, deep layering extraction, image enhancement processing, three-dimensional model construction, and feature matching, effectively solves the problem of low accuracy in identification using traditional single image processing methods in complex industrial environments, significantly improving the accuracy of instrument data identification.

[0028] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any of the foregoing.

[0029] A fourth aspect of this application provides a computer-readable storage medium storing instructions that, when executed, perform the method described in any of the preceding descriptions. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the architecture of an instrument data recognition system disclosed in an embodiment of this application;

[0031] Figure 2 This is a flowchart illustrating an instrument data identification method disclosed in an embodiment of this application;

[0032] Figure 3 yes Figure 2 A flowchart illustrating a sub-step of step S102;

[0033] Figure 4 yes Figure 3 A flowchart illustrating a sub-step of step S202;

[0034] Figure 5 yes Figure 3 A flowchart illustrating a sub-step of step S204;

[0035] Figure 6 yes Figure 2 A flowchart illustrating a sub-step of step S103;

[0036] Figure 7 yes Figure 2 A flowchart illustrating a sub-step of step S104;

[0037] Figure 8 yes Figure 2 A flowchart illustrating a sub-step of step S105;

[0038] Figure 9 This is a schematic diagram of a module of an instrument data recognition system provided in an embodiment of this application;

[0039] Figure 10 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application.

[0040] Figure labeling: 21. Multidimensional data acquisition module; 22. Deep layering processing module; 23. Image enhancement processing module; 24. 3D modeling generation module; 25. Feature matching analysis module; 26. Intelligent detection and retrieval module; 901. Processor; 902. Communication bus; 903. User interface; 904. Network interface; 905. Memory. Detailed Implementation

[0041] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0042] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0043] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0044] Figure 1 An exemplary system architecture 10 for an instrument data recognition system is shown.

[0045] like Figure 1 As shown, system architecture 10 may include electronic device 11, network 12, and server 13. Network 12 serves as the medium for providing a communication link between electronic device 11 and server 13. Network 12 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0046] Operators can use electronic device 11 to interact with server 13 via network 12 to receive or send data, etc. Various identification applications can be installed on electronic device 11, such as instrument data acquisition applications, instrument image processing applications, etc.

[0047] Electronic device 11 is hardware and can be various electronic devices with a display screen, including but not limited to smartphones, tablets, portable testing terminals, and industrial control computers.

[0048] Server 13 can be a backend server providing various identification services, such as a computing server that processes the instrument data collected by electronic device 11. The backend server can analyze and process the received data and feed back the processing results (identification results) to the electronic device.

[0049] The following detailed explanation uses the electronic device side as an example.

[0050] This embodiment discloses a method for identifying instrument data. Figure 2 This is a flowchart illustrating an instrument data identification method disclosed in an embodiment of this application, as shown below. Figure 2 As shown, steps S101 to S106 are included, and the steps are as follows:

[0051] S101: Collect multidimensional image data and structural parameters of the instrument under test. The multidimensional image data includes point cloud data and image data, and the structural parameters include measurement range, scale distribution and display method.

[0052] The instrument to be tested refers to the analog instrument equipment, such as pressure gauges and thermometers, which are measuring instruments with pointers and scales. Multidimensional image data refers to the digital representation data of the instrument acquired from multiple angles and directions. It includes point cloud data and image data. The point cloud data represents the spatial position information of the instrument surface obtained through 3D scanning, consisting of a large number of discrete spatial coordinate points. The image data represents the two-dimensional visible light information of the instrument surface, including features such as color and texture. Structural parameters refer to the basic operating parameters of the instrument. The measurement range indicates the maximum and minimum measurable values ​​of the instrument; the scale distribution indicates the position and interval of the scale lines; and the display method indicates the reading display format, such as circular or fan-shaped.

[0053] Specifically, the electronic device first activates the 3D scanning module and the image acquisition module to perform omnidirectional data acquisition on the instrument under test. The 3D scanning module emits a laser beam to scan the surface, receives reflected signals, and converts them into point cloud data. Simultaneously, the image acquisition module captures high-definition images of the instrument's surface, comprehensively recording the instrument's appearance features and spatial structure information. During the acquisition process, the range information of the instrument's nameplate area is extracted from the acquired image data. Combined with point cloud data analysis, the spacing and distribution characteristics between adjacent scales on the dial are used to obtain the scale distribution pattern. The acquired image data and point cloud data are matched with preset typical features of pointer, digital, and other display methods to identify the display type, and these structural parameters are stored in a database. Through this acquisition process, the electronic device ultimately obtains multi-dimensional image data, including point cloud data and image data, as well as structural parameters containing range, scale distribution, and display method information.

[0054] S102: Perform deep layering processing on the multidimensional image data to extract the instrument layer image data.

[0055] Among them, the instrument layer image data is used to represent the two-dimensional imaging information of the corresponding depth layer on the instrument surface.

[0056] Specifically, the electronic device performs deep layering processing on the acquired multidimensional image data based on spatial positional relationships to remove background influences and accurately acquire instrument surface information. First, the spatial coordinate information of each data point in the point cloud data is used to obtain the distance from each point to the electronic device. Points with close distances are grouped into the same layer, forming multiple positional layers from near to far. After obtaining these positional layers, the electronic device maps each point in the image to the corresponding positional layer. Since the instrument surface forms a relatively complete plane in space, the positional layer of the instrument surface can be identified by recognizing the density and continuity characteristics of points in each layer. Then, the image information corresponding to this layer can be extracted to obtain the instrument layer image data.

[0057] Reference Figure 3 , Figure 3 This is provided by the embodiments of this application. Figure 2 A flowchart illustrating a sub-step of step S102, including steps S201 to S205, is as follows:

[0058] S201: Based on multidimensional image data, construct a depth mapping matrix. The rows of the depth mapping matrix are the vertical coordinates of the image, the columns are the horizontal coordinates of the image, and the matrix elements are the depth values ​​of the image pixels.

[0059] The depth mapping matrix is ​​a data structure that records the depth value of each position in an image in the form of a two-dimensional table. The vertical coordinates of the image represent the position in the up-down direction, the horizontal coordinates represent the position in the left-right direction, and the depth value represents the distance from a certain position in the image to the electronic device.

[0060] Specifically, electronic devices need to organize multidimensional image data into a table format. First, a table is created based on the pixel values ​​of the image data. Then, the spatial position of each point in the point cloud data is mapped to the specific position in the image, and the straight-line distance from this point to the electronic device is used as the depth value. If multiple points overlap at the same position in the image, the value of the closest point is selected and filled into the table. For positions without points in the table, the distance values ​​of the eight surrounding positions are found, and the average of these valid distance values ​​is calculated and filled into the table. After filling all positions, a complete depth mapping matrix is ​​obtained.

[0061] S202: Based on the depth distribution characteristics in the depth mapping matrix, the multidimensional image data is processed into layers to obtain preliminary layering results. The preliminary layering results consist of at least one depth layer, and each depth layer includes an independent 3D point cloud subset generated by depth threshold segmentation.

[0062] Among them, the depth distribution feature represents the distribution pattern of distance values ​​in the table, the preliminary stratification result refers to the result obtained from the first stratification, the depth layer represents the layer where a group of data with similar distances is located, the depth threshold refers to the distance boundary value used to distinguish different layers, and the 3D point cloud subset represents the set of spatial location points within the same layer.

[0063] Specifically, the electronic device counts the frequency of each distance value in the depth mapping matrix and generates a distance distribution histogram. From the histogram, it identifies dense regions of distance values ​​representing the instrument surface. The low points between these dense regions are set as depth thresholds. Then, these depth thresholds are used to divide the multidimensional image data into several groups, with each group of distance values ​​corresponding to a depth layer. At the same time, the point cloud data is assigned to the corresponding depth layer according to the distance values. If the distance value of a point falls within the range of a certain depth layer, the point belongs to that depth layer. After this processing, a preliminary layering result is obtained, which contains at least one depth layer. Each depth layer contains a set of three-dimensional point cloud subsets segmented by the depth thresholds.

[0064] Reference Figure 4 , Figure 4 This is provided by the embodiments of this application. Figure 3 A flowchart illustrating a sub-step of step S202, including steps S301 to S303, is as follows:

[0065] S301: Based on the preset depth threshold and the depth value distribution characteristics in the depth mapping matrix, the data with depth values ​​in the same interval in the multidimensional image data are divided into the same layer to obtain a preliminary hierarchical structure.

[0066] The preset depth threshold represents a pre-defined distance boundary value used to distinguish different levels. The hierarchical structure represents different data levels divided according to distance.

[0067] Specifically, electronic devices pre-set a set of depth thresholds based on experience or actual application requirements. These depth thresholds divide the entire distance range into several intervals. Then, the distribution characteristics of depth values ​​in the depth mapping matrix are analyzed, and the data distribution within each interval is statistically analyzed. Next, all data points in the multidimensional image data whose depth values ​​fall within the same interval are classified into the same layer. In this way, all data points are assigned to different layers according to the interval position of their depth values, ultimately forming a preliminary hierarchical structure containing multiple data layers. Each layer contains all data points whose depth values ​​are within the corresponding interval range.

[0068] S302: Based on the preliminary hierarchical structure, calculate the depth difference between adjacent point cloud data in each layer, and subdivide each layer based on the depth difference to obtain multiple candidate depth layers.

[0069] Among them, the candidate depth layer represents a more refined data layer obtained through subdivision.

[0070] Specifically, after obtaining the initial hierarchical structure, the electronic device needs to further refine the data within each layer. First, it searches for spatially adjacent point cloud data pairs in each layer and calculates the depth difference between these adjacent data points. When it is found that the depth difference between some adjacent points is significantly greater than the depth difference between other adjacent points, it indicates that these points may belong to different object surfaces. Therefore, they need to be divided into different layers. Through this subdivision processing based on depth difference, each of the original layers is further divided into multiple more refined candidate depth layers.

[0071] S303: Based on the preset depth similarity judgment criteria, candidate depth layers are merged to obtain preliminary layering results.

[0072] Among them, the preset depth similarity judgment standard refers to the pre-set judgment rules used to judge whether there is similarity between different levels, and the preliminary stratification result refers to the data stratification result obtained after merging.

[0073] Specifically, the electronic device pre-sets a depth similarity judgment standard according to the actual application requirements. This standard includes parameters such as the threshold range of depth difference and the density distribution of data points within the layer. Then, it calculates the depth difference between adjacent candidate depth layers. When the difference between two candidate depth layers meets the preset similarity judgment standard, the two layers are merged into a new layer. In this way, all candidate depth layers that meet the similarity conditions are gradually merged to finally obtain a preliminary layering result that can reflect the depth distribution characteristics of the scene without being overly subdivided.

[0074] S203: Denoising and fracture repair are performed on each depth layer in the preliminary stratification results to obtain the optimized stratification results.

[0075] Among them, the preliminary stratification result represents the data stratification result obtained from the first stratification, the depth layer represents the layer where a group of data that are close to each other is located, the broken region represents the data region that should be continuous but is discontinuous, and the optimized stratification result represents the more accurate stratified data obtained after denoising and repair.

[0076] Specifically, the electronic device denoises the data within each depth layer of the initial stratification result. It calculates the standard deviation of the depth value differences between all adjacent point pairs within the entire depth layer, setting twice this standard deviation as a preset threshold for outlier detection. Then, it calculates the average difference in depth values ​​between each data point and its eight neighboring points. If the difference between a point's depth value and the average depth value of its surrounding points exceeds this preset threshold, the point is marked as an outlier. For marked outliers, the average depth value of its surrounding normal points can be used to replace the outlier, or it can be directly deleted from the dataset. Next, it detects broken regions in the depth layers. These regions typically appear as discontinuous gaps on what should be continuous object surfaces. For detected broken regions, it determines the depth values ​​of the valid data points at both ends of the broken region. Then, based on these two endpoints, it uses linear interpolation to calculate the depth value of the missing points in the broken region. That is, it calculates the average depth value of the two endpoints as the depth value of the middle position. The calculated depth value is then filled into the corresponding position of the broken region, thus reconnecting broken regions that may belong to the same object surface, resulting in an optimized stratification result.

[0077] S204: Extract the feature sets of each depth layer in the optimized layering result, calculate the similarity between the feature sets of each depth layer and the preset standard feature sets, and select the depth layer with the highest similarity as the instrument layer.

[0078] Among them, the feature set represents a set of parameters used to describe the shape and distribution characteristics of each layer, the preset standard feature set represents the standard parameters set in advance to determine whether it is an instrument layer, and the similarity represents the degree of closeness between two sets of features.

[0079] Specifically, the electronic device identifies the number of data points in each small area to represent the density of the points, detects edges to determine whether the shape is circular or square, and calculates the depth difference between adjacent points to obtain the flatness. At the same time, a standard feature set that the instrument layer should have has been set according to experience or requirements. Then, the absolute value of the difference between the feature set of each layer and the standard features is calculated. The smaller the absolute value, the closer it is to the instrument layer. The layer that is closest is selected as the instrument layer.

[0080] Reference Figure 5 , Figure 5 This is provided by the embodiments of this application. Figure 3 A flowchart illustrating a sub-step of step S204, including steps S401 to S404, is as follows:

[0081] S401: Extract the feature set of each depth layer in the optimized layering result. Each feature set includes geometric shape features and depth distribution features. The depth distribution features are the distribution features obtained by statistically analyzing the depth values.

[0082] Among them, geometric shape features represent the shape and edge features of the layer, depth distribution features represent the distribution pattern of the depth values ​​of data points in the layer, and feature set represents the set of geometric shape features and depth distribution features of a data layer.

[0083] Specifically, the electronic device acquires two types of features for each depth layer: geometric features, which are determined by connecting edge points to obtain a closed contour and then determining its shape type, and by counting the number of points per unit area to obtain the distribution density; and depth distribution features, which are obtained by calculating the average depth of all points in the layer and by calculating the standard deviation of the depth values ​​to obtain the fluctuation range. These two types of features are combined to form the feature set of that layer, and each depth layer obtains a feature set containing both geometric features and depth distribution features.

[0084] S402: Compare each feature in each feature set with the corresponding feature in the standard feature set to determine the initial score for each feature.

[0085] The standard feature set represents a combination of reference features obtained by collecting data from multiple known instrument layers beforehand. The initial score indicates how close a single feature is to the standard features.

[0086] Specifically, the electronic device overlays the actual contour shape with a standard shape. The proportion of the overlapping area to the total area is used as the shape score. The ratio of the actual point density to the standard density is used as the density score. For the comparison of depth distribution features, the difference between the actual average depth and the standard depth is used as the depth score, and the difference between the actual depth fluctuation and the standard fluctuation is used as the fluctuation score. The scores obtained through these calculations are the initial scores for each feature. The higher the score, the closer the feature is to the standard feature.

[0087] S403: Based on the initial score and the corresponding preset weight of each feature, determine the similarity score of each feature set.

[0088] The preset weight represents the importance value of each feature determined through a large number of tests, and the similarity score represents the degree of matching between the feature set as a whole and the standard feature set.

[0089] Specifically, the electronic device acquires preset weights for each feature, such as a weight of 0.4 for shape features, 0.2 for density features, 0.3 for depth features, and 0.1 for fluctuation features. Then, the initial score of each feature is multiplied by the corresponding preset weight, such as a weight of 0.4 for shape, 0.2 for density, 0.3 for depth, and 0.1 for fluctuation. Finally, all weighted scores are summed to obtain the similarity score of the feature set. The higher the similarity score, the more likely the depth layer is to be the instrument layer.

[0090] S404: Select the depth layer with the highest similarity score as the instrument layer.

[0091] Specifically, the electronic device compares the similarity scores of all depth layers, finds the depth layer with the highest score, and identifies it as the instrument layer.

[0092] S205: Convert the instrument layer into a two-dimensional image to obtain the instrument layer image data.

[0093] Specifically, the electronic device identifies the range occupied by the instrument layer in the horizontal and vertical directions, which is used as the width and height of the converted image. Then, each point in the instrument layer is projected from top to bottom onto the corresponding position on the plane. Finally, the depth value of each point is converted into a grayscale value between 0 and 255, thus obtaining a black and white image composed of these grayscale values, which is the instrument layer image data.

[0094] S103: Enhance the image data of the instrument layer to obtain the final image data.

[0095] The final image data represents the enhanced and clearer image data.

[0096] Specifically, the electronic device adjusts the grayscale value of the entire image to the middle range to avoid it being too dark or too bright. Then, it raises the grayscale value that is too high and lowers the grayscale value that is too low to increase the difference in grayscale value, making the contrast between the numbers and scales and the background more obvious. Finally, it strengthens the areas in the image with large changes in grayscale value to make the transition of these edge positions steeper, resulting in final image data that is easier to recognize.

[0097] Reference Figure 6 , Figure 6 This is provided by the embodiments of this application. Figure 2 A flowchart illustrating a sub-step of step S103, including steps S501 to S503, is as follows:

[0098] S501: Perform illumination compensation on the instrument layer image data to obtain the first image data.

[0099] The first image data refers to the image data after illumination compensation.

[0100] Specifically, the electronic device calculates the average gray value of each region in the image, identifies regions whose gray values ​​are significantly higher or lower than the overall average, and then adjusts these regions accordingly. For example, if a region has an average gray value of 180, which is much higher than the overall average of 100, the gray value of that region is reduced to near 100; if a region has an average gray value of 30, which is much lower than the overall average of 100, the gray value of that region is increased to near 100. Through such adjustments, the gray values ​​of different regions of the image are made more similar, resulting in more uniform first image data.

[0101] S502: Perform edge sharpening processing on the first image data to obtain the second image data.

[0102] The second image data represents the image data after edge sharpening.

[0103] Specifically, the electronic device identifies locations in the image where grayscale values ​​change drastically as edge locations, and then increases the difference in grayscale values ​​between adjacent pixels at these edge locations to make the edges clearer, resulting in a second image data with sharper edges.

[0104] S503: Combine the scale distribution to enhance the corresponding regions of the second image data to obtain the final image data.

[0105] The final image data represents the image data after all enhancements have been completed.

[0106] Specifically, electronic devices determine the distribution pattern of scale lines and numbers based on the type of instrument. For example, the scale on a round dial is distributed in a ring, while the scale on a straight watch band is distributed linearly. Then, the contrast of gray values ​​is increased in these scale areas to make the difference between the gray values ​​of the scale and the gray values ​​of the background greater, while maintaining the continuity of the scale lines. Finally, the final image data with clearer and easier-to-read scales and numbers is obtained.

[0107] S104: Based on the final image data and structural parameters, generate a three-dimensional model of the instrument to be tested.

[0108] Among them, the three-dimensional model represents the three-dimensional digital representation of the instrument.

[0109] Specifically, the electronic device establishes the dial structure and scale start and end positions according to the measurement range, then determines the arrangement positions of the scale lines and numbers on the dial according to the scale distribution rules, then establishes a pointer or digital display structure on the dial according to the display method, and finally covers the final image data onto the dial surface through texture mapping to obtain a complete three-dimensional model of the instrument to be tested.

[0110] Reference Figure 7 , Figure 7 This is provided by the embodiments of this application. Figure 2 A flowchart illustrating a sub-step of step S104, including steps S601 to S605, is as follows:

[0111] S601: Construct an initial 3D model based on point cloud data.

[0112] Specifically, the electronic device finds points with similar positions in the spatial point cloud as neighboring points, then connects every three neighboring points with triangles to form a large number of triangular facets, and finally connects all the triangular facets into a whole and performs smoothing to obtain an initial three-dimensional model with a smooth surface.

[0113] S602: Extract the first texture feature from the final image data.

[0114] The first texture feature represents the surface detail features in the image.

[0115] Specifically, electronic devices identify regions in an image where grayscale values ​​change similarly and appear repeatedly. These regions may be patterns or textures on the surface of an instrument, thus obtaining the first texture feature.

[0116] S603: Construct constraints based on the measurement range and display method.

[0117] Specifically, the electronic device determines the size limit of the dial and the range limit of the scale area based on the range, and then determines whether to follow the circular structure of the pointer dial or the rectangular structure of the digital dial based on the display method, and finally obtains the constraint conditions.

[0118] S604: Register the initial 3D model with the base model with the highest matching degree in the preset instrument base template library to obtain the registered 3D model.

[0119] Specifically, the electronic device registers the initial 3D model with the models in the base template library. First, it compares the initial 3D model with all the models in the base template library to find the most similar base model. Then, it adjusts the position and orientation of the initial 3D model to make it have the highest degree of overlap with the selected base model. Finally, it obtains a registered 3D model that matches the standard model.

[0120] S605: Optimize the registered 3D model by combining texture features and constraints to obtain the 3D model.

[0121] The three-dimensional model represents the optimized stereoscopic digital representation.

[0122] Specifically, the electronic device adds texture features to the surface of the registered 3D model to give it realistic appearance details, and then adjusts the shape and size of the model according to the constraints to meet the requirements of the measurement range and display method, finally obtaining the 3D model.

[0123] S105: Extract the key features of the 3D model and match them with the standard features in the preset instrument library to obtain the matching results.

[0124] Among them, key features represent the main characteristic parameters of the instrument, preset instrument library represents the information of multiple standard instruments stored, standard features represent the characteristic parameters recorded in the preset instrument library, and matching results represent the comparison results of features.

[0125] Specifically, the electronic device obtains feature parameters such as the shape and size of the dial, the distribution and spacing of the scale from the 3D model, and combines these parameters into a set of feature data. Then, it subtracts the absolute value of each standard feature data in the preset instrument library from this set of data to obtain the error value. The similarity is then calculated based on the magnitude of the error value. The smaller the error value, the higher the similarity. Finally, the similarity with each standard feature is obtained as the matching result.

[0126] Reference Figure 8 , Figure 8 This is provided by the embodiments of this application. Figure 2 A flowchart illustrating a sub-step of step S105, including steps S701 to S703, is as follows:

[0127] S701: Extracts the geometric features, secondary texture features, and structural features of the 3D model.

[0128] Among them, geometric features represent the shape and size information of the object, including basic dimensions such as length, width, and height, and shape features such as circle or square. Secondary texture features represent the scale and indication information on the instrument surface, including the distribution of scale lines, numerical markings, pointer position, etc. Structural features represent the connection area, fit method, clearance size, and other fitting parameters obtained by statistically analyzing the contact area of ​​the component surface.

[0129] Specifically, the electronic device first acquires the coordinates of all grid points on the surface of the 3D model, calculates the distance between adjacent points to obtain the side length data, calculates geometric features such as dial diameter and casing depth based on the side length data, then calculates the height difference between each grid point and its surrounding adjacent points to obtain surface morphology features, identifies areas with large height differences as locations of concavity and convexity changes, then analyzes the surface image to obtain texture features, marks the start and end coordinates of the scale lines, the center coordinates of the numerals, and the rotation angle of the pointer, and finally counts the common surface areas between components to obtain structural features, and records assembly parameters such as the contact area and installation gap between the dial and casing, and the positioning area and fitting gap between the pointer and the shaft, thereby completing the extraction of 3D model features.

[0130] S702: Use geometric and structural features to perform matching based on the corresponding standard geometric and structural features in the preset instrument library to obtain the first matching result.

[0131] Among them, the preset instrument library represents a dataset that stores multiple standard instruments and their features; the standard geometric features represent the basic data such as the size and shape parameters of each standard instrument in the instrument library; the standard structural features represent the assembly data such as the assembly parameters and connection relationships of each standard instrument in the instrument library; and the first matching result represents the standard instrument type with the smallest feature difference.

[0132] Specifically, the electronic device reads the feature data of all standard instruments in the preset instrument library, calculates the feature differences between each standard instrument and the instrument to be identified, and calculates the dimensional difference value by comparing the geometric features of the instrument to be identified, such as the dial diameter and shell depth, with the standard geometric features. It also calculates the assembly difference value by comparing the structural features of the instrument to be identified, such as the contact area and fit clearance, with the standard structural features. Then, the geometric difference value and the structural difference value are summed to obtain the total difference value. Finally, the instrument type with the smallest total difference value is selected as the first matching result.

[0133] S703: Match the second texture feature with the corresponding standard texture feature in the first matching result to obtain the matching result.

[0134] Specifically, the electronic device compares the texture features of the standard instrument in the first matching result, counts whether the number of scale lines of the instrument to be identified is the same, whether the distribution spacing is consistent, counts whether the number of digital labels is the same, whether the displayed content is consistent, and confirms whether the current angle of the pointer is within the standard movement range. If the matching degree of the above three features is higher than the preset threshold, it is determined to be the same model of instrument, thus obtaining the final matching result.

[0135] S106: Based on the matching results, call the detection model corresponding to the instrument under test to perform data recognition on the instrument under test.

[0136] Specifically, the electronic device determines the specific model of the instrument to be tested based on the matching results, and calls the corresponding detection model from the model library. This detection model contains identification parameters such as the scale spacing ratio, numerical mapping relationship, and pointer reading rules of this type of instrument. Then, based on the current pointer angle of the instrument to be tested, combined with the scale distribution and numerical mapping relationship, the pointer angle position is converted into the actual display value, thereby completing the identification of the instrument data.

[0137] This embodiment also discloses an instrument data recognition system 20, referencing... Figure 9 Specifically, it includes:

[0138] The multidimensional data acquisition module 21 is used to acquire multidimensional image data and structural parameters of the instrument under test. The multidimensional image data includes point cloud data and image data, and the structural parameters include measurement range, scale distribution and display mode.

[0139] The depth layering processing module 22 is used to perform depth layering processing on the multidimensional image data and extract the instrument layer image data;

[0140] Image enhancement processing module 23 is used to enhance the image data of the instrument layer to obtain the final image data;

[0141] The 3D modeling generation module 24 is used to generate a 3D model of the instrument to be tested based on the final image data and the structural parameters.

[0142] The feature matching and analysis module 25 is used to extract the key features of the three-dimensional model and match them with the standard features in the preset instrument library to obtain the matching results;

[0143] The intelligent detection invocation module 26 is used to invoke the detection model corresponding to the instrument under test to perform data identification on the instrument under test based on the matching result.

[0144] Optionally, the depth layering processing module 22 is further configured to construct a depth mapping matrix based on the multidimensional image data, wherein the rows of the depth mapping matrix are the vertical coordinates of the image, the columns are the horizontal coordinates of the image, and the matrix elements are the depth values ​​of image pixels; perform layering processing on the multidimensional image data according to the depth distribution characteristics in the depth mapping matrix to obtain a preliminary layering result, wherein the preliminary layering result consists of at least one depth layer, and each depth layer includes an independent 3D point cloud subset generated by depth threshold segmentation; perform denoising and fracture region repair processing on each depth layer in the preliminary layering result to obtain an optimized layering result; extract the feature set of each depth layer in the optimized layering result, calculate the similarity between the feature set of each depth layer and a preset standard feature set, and select the depth layer with the highest similarity as the instrument layer; convert the instrument layer into a 2D image to obtain instrument layer image data.

[0145] Optionally, the depth layering processing module 22 is further configured to divide data in the multidimensional image data whose depth values ​​are in the same interval into the same layer according to a preset depth threshold and the depth value distribution characteristics in the depth mapping matrix, to obtain a preliminary hierarchical structure; based on the preliminary hierarchical structure, calculate the depth difference between adjacent point cloud data in each layer, subdivide each layer based on the depth difference, to obtain multiple candidate depth layers; and merge the candidate depth layers based on a preset depth similarity judgment criterion to obtain a preliminary layering result.

[0146] Optionally, the depth layering processing module 22 is further configured to extract feature sets of each depth layer in the optimized layering result, wherein each feature set includes geometric shape features and depth distribution features, wherein the depth distribution features are distribution features obtained by statistical analysis of depth values; compare each feature in each feature set with the corresponding features in the standard feature set to determine the initial score corresponding to each feature; determine the similarity score of each feature set based on the initial score and the corresponding preset weight; and select the depth layer with the highest similarity score as the instrument layer.

[0147] Optionally, the image enhancement processing module 23 is further configured to perform illumination compensation on the instrument layer image data to obtain first image data; perform edge sharpening processing on the first image data to obtain second image data; and perform corresponding region enhancement on the second image data in conjunction with the scale distribution to obtain final image data.

[0148] Optionally, the 3D modeling generation module 24 is further configured to: construct an initial 3D model based on the point cloud data; extract the first texture feature from the final image data; construct constraints according to the measurement range and the display method; register the initial 3D model with the base model with the highest matching degree in the preset instrument base template library to obtain a registered 3D model; and optimize the registered 3D model by combining the texture feature and the constraints to obtain a 3D model.

[0149] Optionally, the feature matching analysis module 25 is further used to extract the geometric features, second texture features, and structural features of the three-dimensional model; to match the geometric features and structural features with the corresponding standard geometric features and standard structural features in a preset instrument library to obtain a first matching result; and to match the second texture features with the corresponding standard texture features in the first matching result to obtain a matching result.

[0150] It should be noted that the above embodiments of the apparatus are only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0151] This embodiment also discloses an electronic device 900, as shown in the reference... Figure 10 The electronic device may include: at least one processor 901, at least one communication bus 902, user interface 903, network interface 904, and at least one memory 905.

[0152] The communication bus 902 is used to enable communication between these components.

[0153] The user interface 903 may include a display screen and a camera. Optional user interfaces may also include standard wired interfaces and wireless interfaces.

[0154] The network interface 904 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0155] The processor 901 may include one or more processing cores. The processor connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory, and by calling data stored in memory. Optionally, the processor may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also be implemented as a separate chip without being integrated into the processor.

[0156] The memory 905 may include random access memory (RAM) or read-only memory. Optionally, the memory may include a non-transitory computer-readable storage medium. The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor. As shown in the figure, the memory, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for an instrument data recognition method.

[0157] exist Figure 10In the electronic device shown, the user interface is mainly used to provide an input interface for the user and to obtain the user input data; while the processor can be used to call an application program that stores an instrument data recognition method in the memory. When executed by one or more processors, the electronic device performs one or more methods as described in the above embodiments.

[0158] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0159] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0160] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some service interfaces; indirect couplings or communication connections between apparatuses or units may be electrical or other forms.

[0161] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0162] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0163] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0164] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Other embodiments of this disclosure will be readily apparent to those skilled in the art upon consideration of the disclosure in this specification. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for identifying instrument data, characterized in that, The method includes: Collect multidimensional image data and structural parameters of the instrument under test. The multidimensional image data includes point cloud data and image data. The structural parameters include measurement range, scale distribution and display method. Based on the multidimensional image data, a depth mapping matrix is ​​constructed. The rows of the depth mapping matrix are the vertical coordinates of the image, the columns are the horizontal coordinates of the image, and the matrix elements are the depth values ​​of the image pixels. Based on the depth distribution features in the depth mapping matrix, the multidimensional image data is processed into layers to obtain a preliminary layering result. The preliminary layering result consists of at least one depth layer, and each depth layer includes an independent 3D point cloud subset generated by depth threshold segmentation. The preliminary layering results are processed by denoising and fracture repair to obtain optimized layering results. The feature sets of each depth layer in the optimized layering results are extracted, the similarity between the feature sets of each depth layer and the preset standard feature sets is calculated, and the depth layer with the highest similarity is selected as the instrument layer. The instrument layer is converted into a two-dimensional image to obtain instrument layer image data; The image data of the instrument layer is enhanced to obtain the final image data; Based on the final image data and the structural parameters, a three-dimensional model of the instrument to be tested is generated; Key features of the 3D model are extracted and matched with standard features in a preset instrument library to obtain matching results; Based on the matching result, the detection model corresponding to the instrument under test is invoked to perform data identification on the instrument under test.

2. The method according to claim 1, characterized in that, The multidimensional image data is layered based on the depth distribution features in the depth mapping matrix to obtain a preliminary layering result. This preliminary layering result consists of at least one depth layer, which is an independent subset of a 3D point cloud generated through depth thresholding. Specifically, it includes: Based on the preset depth threshold and the depth value distribution characteristics in the depth mapping matrix, the data in the multidimensional image data whose depth values ​​are in the same interval are divided into the same layer to obtain a preliminary hierarchical structure. Based on the preliminary hierarchical structure, the depth difference between adjacent point cloud data in each layer is calculated, and each layer is subdivided based on the depth difference to obtain multiple candidate depth layers. Based on the preset depth similarity judgment criteria, the candidate depth layers are merged to obtain preliminary layering results.

3. The method according to claim 1, characterized in that, The process of extracting feature sets from each depth layer in the optimized layering result, calculating the similarity between the feature sets of each depth layer and the standard feature set, and selecting the depth layer with the highest similarity as the instrument layer specifically includes: Extract the feature set of each depth layer in the optimized layering result. Each feature set includes geometric shape features and depth distribution features. The depth distribution features are distribution features obtained by statistically analyzing depth values. Each feature in each feature set is compared with the corresponding feature in the standard feature set to determine the initial score corresponding to each feature; Based on the initial score and the corresponding preset weight of each feature, the similarity score of each feature set is determined; The depth layer with the highest similarity score is selected as the instrument layer.

4. The method according to claim 1, characterized in that, The enhancement processing of the instrument layer image data to obtain the final image data specifically includes: Illumination compensation is applied to the instrument layer image data to obtain the first image data; The first image data is subjected to edge sharpening processing to obtain the second image data; The second image data is enhanced in the corresponding regions based on the scale distribution to obtain the final image data.

5. The method according to claim 1, characterized in that, The step of generating a three-dimensional model of the instrument to be tested based on the final image data and the structural parameters specifically includes: An initial 3D model is constructed based on the point cloud data; Extract the first texture feature from the final image data; Constraints are constructed based on the measurement range and the display method; The initial 3D model is registered with the basic model with the highest matching degree in the preset instrument basic template library to obtain the registered 3D model; The registered 3D model is optimized by combining the texture features and the constraints to obtain a 3D model.

6. The method according to claim 1, characterized in that, The process of extracting key features from the 3D model and matching them with standard features in a preset instrument library to obtain matching results specifically includes: Extract the geometric features, second texture features, and structural features of the three-dimensional model; The geometric features and structural features are used to perform matching based on the corresponding standard geometric features and standard structural features in the preset instrument library to obtain a first matching result; The second texture feature is matched with the corresponding standard texture feature in the first matching result to obtain the matching result.

7. An instrument data recognition system, characterized in that, For implementing the instrument data identification method as described in claim 1, the instrument data identification system comprises: The multidimensional data acquisition module is used to acquire multidimensional image data and structural parameters of the instrument under test. The multidimensional image data includes point cloud data and image data, and the structural parameters include measurement range, scale distribution and display method. The deep layering processing module is used to perform deep layering processing on the multidimensional image data and extract the instrument layer image data; An image enhancement processing module is used to enhance the image data of the instrument layer to obtain the final image data; A 3D modeling generation module is used to generate a 3D model of the instrument to be tested based on the final image data and the structural parameters. The feature matching and analysis module is used to extract key features of the three-dimensional model and match them with standard features in a preset instrument library to obtain matching results; The intelligent detection invocation module is used to invoke the detection model corresponding to the instrument under test to perform data identification on the instrument under test based on the matching result.

8. An electronic device, characterized in that, The device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed, perform the method as described in any one of claims 1-6.

Images