A visual data acquisition system and method in a water meter verification process

Abstract

The application discloses a visual data acquisition system and method in a water meter verification process, and relates to the field of water meter production and detection.The system comprises a matrix sensing module, which is used for sensing water meter surface distance information and constructing a water meter surface heterogeneous model based on the water meter surface distance information; and a modeling module, which is used for constructing a water meter surface heterogeneous sample model.The application combines a ranging matrix to dynamically debug the camera posture for collecting water meter surface image data, so that the water meter surface image data collected each time is aligned with the constructed water meter surface heterogeneous model and water meter surface heterogeneous sample model in the collection stage, thereby improving the consistency of the water meter surface image data, and further combining image segmentation and specific image enhancement processing logic to output water meter verification image data, so that the water meter verification system can perform a verification operation with high-precision water meter surface image.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water meter production detection, in particular to a visual data acquisition system and method in the water meter calibration process. BACKGROUND

[0002] Water meter production appearance detection mainly checks whether the shell, dial and other components have cracks, deformation, rust, whether the coating is uniform and smooth, whether the scale and numbers are clear and wear-resistant, whether the interface thread is regular, whether the lead seal device is intact, and whether the appearance quality meets the standard through visual, touch and special tool measurement to avoid affecting the measurement accuracy and use safety.

[0003] In the prior art, water meter production appearance detection is often assisted by a machine vision system, which collects water meter images through a camera, detects surface scratches, stains, missing parts or misinstallation of parts, etc. by using image processing algorithms, and some systems also combine deep learning models to improve the recognition rate of complex defects. In addition, laser scanning technology is also used to detect whether the appearance contour meets the standard, and manual sampling inspection is also used to ensure the detection accuracy, so as to realize automatic or semi-automatic detection of water meter appearance size, surface quality and other multi-dimensional, improve production efficiency and product quality consistency.

[0004] However, in the process of water meter appearance detection, the accuracy of water meter surface image data collection directly affects the water meter appearance detection result, and how to accurately collect water meter surface image data still needs new collection technology to solve this technical problem.

[0005] Therefore, a visual data acquisition system and method in the water meter calibration process are provided. SUMMARY

[0006] In view of the above shortcomings of the prior art, the present application provides a visual data acquisition system and method in the water meter calibration process, which can effectively solve the problems of the prior art.

[0007] To achieve the above purpose, the present application is realized by the following technical scheme.

[0008] The present application discloses a visual data acquisition system in the water meter calibration process, which comprises:

[0009] The matrix perception module is used for perceiving water meter surface distance information and constructing a water meter surface heterogeneous model based on the water meter surface distance information; the modeling module is used for constructing a water meter surface heterogeneous sample model; the analysis module is used for analyzing the similarity between the water meter surface heterogeneous model constructed by the matrix perception module in real time and the water meter surface heterogeneous sample model constructed by the modeling module, and performing a collection operation of water meter surface image data based on the similarity analysis result; the selection module is used for traversing the water meter surface image data collected by the analysis module, selecting one water meter surface image data from the water meter surface image data as image data for water meter calibration; the preprocessing module is used for receiving the water meter surface image data as the image data for water meter calibration in the selection module, and performing segmentation and enhancement processing on the image data; and the output module is used for acquiring the water meter surface image after the enhancement processing, and transmitting the water meter surface image after the enhancement processing to a water meter calibration system connected to the system.

[0010] Further, the matrix perception module is integrated by a plurality of micro distance measuring sensors, the plurality of micro distance measuring sensors are arranged at equal distances in the horizontal direction and the vertical direction, and the distance measuring ends of the micro distance measuring sensors are all located on the same plane.

[0011] The water meter is output by a water meter production device, and the matrix perception module is triggered to operate when the water meter is conveyed by an output conveyor to be directly below the matrix perception module.

[0012] The matrix perception module is connected to a mechanical arm, and the mechanical arm controls the matrix perception module to move horizontally in a spiral manner in the triggered operation state of the matrix perception module, perceives the water meter surface distance information based on a specified frequency during the horizontal spiral movement, the specified frequency is not greater than 10 times per second, the horizontal spiral movement speed is not greater than 1 mm / s, and the maximum diameter of the horizontal spiral movement path is not greater than 3 mm.

[0013] Further, the matrix perception module is integrated by a plurality of micro distance measuring sensors, the plurality of micro distance measuring sensors are arranged at equal distances in the horizontal direction and the vertical direction, and the distance measuring ends of the micro distance measuring sensors are all located on the same plane.

[0014] The distance measuring results of the micro distance measuring sensors are obtained, in the three-dimensional space, a longitudinal line segment representing each distance measuring result is drawn based on a plane, the bottom ends of the drawn line segments are located on the same plane, the arrangement postures of the micro distance measuring sensors are consistent with the bottom ends, the distances from the top ends of the drawn line segments to the bottom ends are consistent with the corresponding distance measuring results, and a closed three-dimensional figure composed of a plurality of triangular faces is obtained by connecting the adjacent vertices of the line segments, and is recorded as a water meter surface heterogeneous model.

[0015] Further, the modeling module runs in a stage, a sample water meter is placed on the output conveyor belt by a system end user, and the sample water meter is directly below the matrix sensing module, the matrix sensing module runs to sense the surface distance information of the sample water meter, and a water meter surface heterogeneous sample model is constructed based on the distance information.

[0016] Further, in the micro distance measuring sensor of the matrix sensing module, a micro high-definition industrial camera is disposed in the interval space of the center position micro distance measuring sensor, and the high-definition industrial camera is used to collect water meter surface image data;

[0017] The micro high-definition industrial camera is synchronously run with each running of the matrix sensing module, and is subject to:

[0018] In the first running of the matrix sensing module, the micro high-definition industrial camera is synchronously run to collect a water meter surface image data, the analysis module is synchronously run with each running of the matrix sensing module, and similarity analysis is continuously performed, and the similarity analysis result is recorded from the second running of the matrix sensing module;

[0019] In each similarity analysis result is greater than the last similarity analysis result, the micro high-definition industrial camera is run once to perform water meter surface image data collection, and each collected water meter surface image data is marked with its corresponding similarity analysis result; in each similarity analysis result is less than or equal to the last similarity analysis result, the water meter surface image data collection operation is not performed;

[0020] The analysis module is internally provided with a storage unit, and the storage unit is used to store the water meter surface image data collected by the micro high-definition industrial camera.

[0021] Further, the similarity analysis logic of the water meter surface heterogeneous model and the water meter surface heterogeneous sample model in the analysis module is represented as:

[0022] ;

[0023] In the formula: is the similarity of the water meter surface heterogeneous model a and the water meter surface heterogeneous sample model b; is the total amount of the vertex control group in the water meter surface heterogeneous model a and the water meter surface heterogeneous sample model b; is the offset distance of the vertex from the water meter surface heterogeneous model a relative to the vertex from the water meter surface heterogeneous sample model b in the i-th control group; is the longest side of the bottom surface of the local model in which the vertex from the water meter surface heterogeneous model a is located, and the longest side of the bottom surface of the local model in which the vertex from the water meter surface heterogeneous model b is located in the i-th control group;

[0024] Wherein, each vertex in each model, that is, the vertex of each drawing line segment, is composed of two model vertices, and the two vertices are derived from the water surface isomorphic model a and the water surface isomorphic sample model b, and the two vertices in one vertex pair correspond to the same miniature distance measuring sensor, and the local model where the vertex is located is the water surface isomorphic model a and the water surface isomorphic sample model b, and the side where each vertex is located is a triangular local solid model.

[0025] Further, the selection module takes the water surface image data stored in the storage unit as the traversal target group when traversing the water surface image data, performs the traversal operation, and selects the water surface image data based on the similarity analysis result marked by each water surface image data, so that the water surface image data with the highest similarity analysis result is selected, and the remaining water surface image data is deleted as the deletion target in the storage unit.

[0026] Further, the system is preset to be connected to any existing water meter calibration system for water meter appearance calibration through water surface images.

[0027] Further, the matrix sensing module and the modeling module are interactively connected with the analysis module through a wireless network, the analysis module is interactively connected with the storage unit through a wireless network, the analysis module is interactively connected with the selection module through a wireless network, the selection module is interactively connected with the storage unit through a wireless network, and the selection module is interactively connected with the preprocessing module and the output module through a wireless network.

[0028] On the other hand, a visual data acquisition method in a water meter calibration process comprises the following steps:

[0029] A distance measuring matrix is constructed, the distance measuring matrix is used to sense the water surface distance information, and a water surface isomorphic model is constructed based on the sensed water surface distance information; a water surface isomorphic sample model is constructed by combining a sample water meter with the distance measuring matrix, water surface image data is acquired in real time based on the similarity analysis of the water surface isomorphic sample model and the water surface isomorphic model; a similarity analysis result is marked for each acquired water surface image data, one water surface image data is selected based on the similarity analysis result marked by each acquired water surface image data, and the selected water surface image data is used as image data for water meter calibration; the water surface image data used as the image data for water meter calibration is segmented and enhanced to obtain a water surface image; and the obtained water surface image is forwarded to a preset water meter calibration system, and subsequent water meter appearance calibration is completed through the water meter calibration system.

[0030] Compared with the known prior art, the technical scheme provided by the application has the following beneficial effects:

[0031] The application provides a visual data acquisition system and method in a water meter calibration process. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0033] Figure 1 It is a structural schematic diagram of a visual data acquisition system in a water meter calibration process.

[0034] Figure 2 It is a flowchart of a visual data acquisition method in a water meter calibration process.

[0035] Figure 3 It is an example schematic diagram of the running posture of the matrix perception module in the present application. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0037] The present application will be further described below in combination with the embodiments.

[0038] Embodiment 1:

[0039] A visual data acquisition system in a water meter calibration process in the present embodiment, as shown in Figure 1 , comprises:

[0040] A matrix sensing module is configured to sense water meter surface distance information and construct a water meter surface heterogeneous model based on the water meter surface distance information.

[0041] The matrix sensing module is integrated with a plurality of miniature distance measuring sensors, which are arranged at equal distances in the horizontal and vertical directions, and the distance measuring ends of the miniature distance measuring sensors are located on the same plane.

[0042] The water meter is output by a water meter production device, and when the water meter is delivered by an output conveyor to directly below the matrix sensing module, the matrix sensing module is triggered to operate.

[0043] The matrix sensing module is connected to a mechanical arm, which controls the horizontal spiral movement of the matrix sensing module in the triggered operating state of the matrix sensing module, and based on a specified frequency, the water meter surface distance information is sensed during the horizontal spiral movement. The specified frequency is not greater than 10 times per second, the horizontal spiral movement speed is not greater than 1 mm per second, and the maximum diameter of the horizontal spiral movement path is not more than 3 mm.

[0044] In the operating stage of the matrix sensing module, the construction of the water meter surface heterogeneous model is performed synchronously once for each sensing of the water meter surface distance information.

[0045] The distance measuring results of the miniature distance measuring sensors are obtained, and in the three-dimensional space, a longitudinal line segment representing each distance measuring result is drawn based on a plane, so that the bottom ends of the drawn line segments are located on the same plane, and the bottom ends are consistent with the arrangement posture of the miniature distance measuring sensors. The distance from the top end of each drawn line segment to the bottom end is consistent with the corresponding distance measuring result. Based on the adjacent connection of the vertices of each line segment, a closed three-dimensional figure composed of a plurality of triangular faces is obtained, which is denoted as a water meter surface heterogeneous model.

[0046] A modeling module is configured to construct a water meter surface heterogeneous sample model.

[0047] In the operating stage of the modeling module, a sample water meter is placed on the output conveyor by a system user, and the sample water meter is directly below the matrix sensing module. The matrix sensing module senses the sample water meter surface distance information and constructs a water meter surface heterogeneous sample model based on the distance information.

[0048] An analysis module is configured to analyze the similarity between the water meter surface heterogeneous model constructed by the real-time operation of the matrix sensing module and the water meter surface heterogeneous sample model constructed by the operation of the modeling module. Based on the similarity analysis result, the acquisition operation of the water meter surface image data is performed.

[0049] In the miniature distance measuring sensors of the matrix sensing module, a miniature high-definition industrial camera is disposed in the space between the center position miniature distance measuring sensors, and the water meter surface image data is acquired by the high-definition industrial camera.

[0050] The miniature high-definition industrial camera runs synchronously with each running of the matrix perception module, and is subject to:

[0051] The matrix perception module runs for the first time, the miniature high-definition industrial camera runs synchronously to collect a water meter surface image data, the analysis module runs synchronously with each running of the matrix perception module, continuously performs similarity analysis, and records the similarity analysis result since the second running of the matrix perception module;

[0052] In each similarity analysis result is greater than the last similarity analysis result, the miniature high-definition industrial camera runs once to perform water meter surface image data collection, and each collected water meter surface image data is marked with its corresponding similarity analysis result; in each similarity analysis result is less than or equal to the last similarity analysis result, no water meter surface image data collection operation is performed;

[0053] The analysis module is internally provided with a storage unit, and the storage unit is used to store the water meter surface image data collected by the miniature high-definition industrial camera;

[0054] The similarity analysis logic of the water meter surface heterogeneous model and the water meter surface heterogeneous sample model in the analysis module is represented as:

[0055] ;

[0056] In the formula: is the similarity of the water meter surface heterogeneous model a and the water meter surface heterogeneous sample model b; is the total amount of the vertex comparison group of the water meter surface heterogeneous model a and the water meter surface heterogeneous sample model b; is the offset distance of the vertex originating from the water meter surface heterogeneous model a relative to the vertex originating from the water meter surface heterogeneous sample model b in the i-th comparison group; is the longest side of the bottom surface of the local model in which the vertex originating from the water meter surface heterogeneous model a is located, and the longest side of the bottom surface of the local model in which the vertex originating from the water meter surface heterogeneous model b is located in the i-th comparison group;

[0057] In the formula, each vertex in each model, i.e., the vertex of each drawn line segment, is composed of two model vertices, and the two vertices originate from the water meter surface heterogeneous model a and the water meter surface heterogeneous sample model b, respectively, and the two vertices in one vertex comparison group correspond to the same miniature distance measuring sensor, and the local model in which the vertex is located is a local three-dimensional model in which each vertex is located on a triangular side surface of the water meter surface heterogeneous model a and the water meter surface heterogeneous sample model b;

[0058] Through the above logical formula, the offset distance of the corresponding vertices in the water meter surface heterogeneous model and the water meter surface heterogeneous sample model is used to measure the difference between the two models, and the parameter The two models are normalized to limit the difference between the sum of the two models to the range of 0-1, and then 1 is subtracted from the normalized and averaged value to obtain the similarity of the two models, which can well reflect the similarity of the two models, thereby providing logical support for the selection of subsequent water meter surface image data;

[0059] The selection module selects a water meter surface image data from the water meter surface image data as the image data for water meter calibration;

[0060] When the selection module traverses the water meter image data, the water meter surface image data stored in the storage unit is used as the traversal target group, and the traversal operation is performed. Based on the similarity analysis result marked by each water meter surface image data, the water meter surface image data with the highest similarity analysis result is selected, and the remaining water meter surface image data is deleted from the storage unit.

[0061] The preprocessing module receives the water meter surface image data selected as the image data for water meter calibration in the selection module, and performs segmentation and enhancement processing on the image data;

[0062] The segmentation operation of the water meter surface image data in the preprocessing module is subject to:

[0063] In the system configuration and deployment phase, the color of the conveyor belt outputting the water meter on the water meter production equipment is distinguished from the color of each component of the water meter, so that the pixel-level color pixel block is distinguished to obtain a pixel set representing the water meter surface image. Further, based on the pixel set, a segmentation operation is performed on the water meter surface image data to segment the image of the water meter surface region from the water meter surface image data, i.e., the water meter surface image. Further, the water meter surface image is enhanced.

[0064] The enhancement processing operation of the water meter surface image in the preprocessing module is subject to:

[0065] ;

[0066] In the formula: is the enhanced image; is the local adaptive coefficient; is the gray scale transformation coefficient; is the original water meter surface image; is the power transformation exponent; is the filter fusion coefficient; indicates that the original water meter surface image is subjected to Gaussian filtering; is the offset;

[0067] It should be noted that:

[0068] ;

[0069] In the formula: is the variance of the local region of the image; is a preset minimum variance threshold value; is a constant, in the range of 0-1;

[0070] is the local mean of the image; is the global maximum mean of the image; is a very small positive number, and the default value is 0.1;

[0071] is the contrast of the local region; is the global average contrast; is an adjustment factor;

[0072] is the edge strength of the image; is the maximum edge strength value in the image;

[0073] is an adjustment coefficient; is the gray level histogram of the local region I(x, y); is the mean of the gray level histogram; is a preset target average brightness value;

[0074] The formula can dynamically adjust the enhancement strategy according to the local features of the image through multiple adaptive parameters, adapt to water meter images under different lighting conditions, noise levels and image contents, and effectively suppress noise while enhancing image details by combining nonlinear gray scale transformation, Gaussian filtering and adaptive adjustment, thereby avoiding the problems of over-enhancement or detail loss that may occur in traditional methods. In addition, the formula focuses on improving the clarity and recognizability of key information such as dial scales and numbers, and provides high-quality image data for subsequent image recognition and reading analysis.

[0075] The output module is configured to obtain the enhanced water meter surface image and transmit the enhanced water meter surface image to a water meter calibration system connected to the system;

[0076] The water meter calibration system connected to the system is any existing system for calibrating the appearance of a water meter through a water meter surface image;

[0077] The matrix perception module and the modeling module are interactively connected with the analysis module through a wireless network, the analysis module is interactively connected with a storage unit through a wireless network, the analysis module is interactively connected with the selection module through a wireless network, the selection module is interactively connected with the storage unit through a wireless network, and the selection module is interactively connected with the preprocessing module and the output module through a wireless network.

[0078] In the embodiment, the matrix perception module runs to perceive the water meter surface distance information, constructs a water meter surface heterogeneous model based on the water meter surface distance information, the modeling module synchronously constructs a water meter surface heterogeneous sample model, the analysis module runs to analyze the similarity between the water meter surface heterogeneous model constructed by the matrix perception module in real time and the water meter surface heterogeneous sample model constructed by the modeling module, based on the similarity analysis result, performs the collection operation of the water meter surface image data, the storage unit synchronously stores the water meter surface image data collected by the miniature high-definition industrial camera, and then the selection module traverses the water meter surface image data collected by the analysis module, selects one piece of water meter surface image data as the image data for water meter calibration in the water meter surface image data, and receives the water meter surface image data as the image data for water meter calibration in the selection module. The image data is segmented and enhanced by the preprocessing module, and finally the enhanced water meter surface image is obtained through the output module. The enhanced water meter surface image is transmitted to the water meter calibration system connected by the system.

[0079] It should be noted that the water meter calibration system mentioned by the output module is preset by the system end user, and the water meter calibration system calibrates whether the appearance of the water meter is qualified through water meter surface image analysis, for example: a water meter detection system based on YOLOv5, a water meter reading recognition system based on CRNN.

[0080] In the above embodiment, the system constructs a heterogeneous model by perceiving the water meter surface distance information, and collects images by similarity analysis with a sample model, which can accurately capture the water meter surface image with the highest reference value. In the collection process, the image is dynamically selected and preprocessed according to the similarity analysis result to ensure that high-quality calibration images are obtained. This method realizes intelligent and accurate collection of water meter surface images, improves the accuracy and effectiveness of image collection compared with traditional collection methods, provides more reliable data support for water meter appearance calibration, and improves the quality and precision of water meter calibration results.

[0081] Referring to Figure 3 As shown in the figure, the upper spiral line of the figure shows the horizontal spiral movement form of the matrix perception module controlled by the mechanical arm, and the lower arrow matrix of the figure shows the distribution posture of the miniature ranging sensor in the matrix perception module and vice versa.

[0082] Embodiment 2:

[0083] In a specific implementation level, on the basis of Embodiment 1, the present embodiment refers to Figure 2 The visual data acquisition system in a water meter calibration process in Embodiment 1 is further specifically described:

[0084] A visual data acquisition method in a water meter calibration process, comprising the following steps:

[0085] Step 1: Construct a ranging matrix, apply the ranging matrix to perceive the distance information of the water meter surface, and construct a water meter surface heterogeneous model based on the perceived distance information of the water meter surface;

[0086] Step 2: Construct a water meter surface heterogeneous sample model by combining the ranging matrix with a sample water meter, and acquire water meter surface image data in real time based on the similarity analysis of the water meter surface heterogeneous sample model and the water meter surface heterogeneous model;

[0087] Step 3: Label the similarity analysis result for each acquired water meter surface image data, and select a water meter surface image data based on the labeled similarity analysis result of each acquired water meter surface image data, and take the selected water meter surface image data as image data for water meter calibration;

[0088] Step 4: Segment and enhance the water meter surface image data as image data for water meter calibration to obtain a water meter surface image;

[0089] Step 5: Forward the obtained water meter surface image to a preset water meter calibration system, and complete subsequent water meter appearance calibration work through the water meter calibration system.

[0090] In summary, in the above-mentioned embodiments, the system and method dynamically debug the camera pose for acquiring water meter surface image data in combination with the ranging matrix during the execution process, so that each time the water meter surface image data is collected, it is aligned with the constructed water meter surface heterogeneous model and water meter surface heterogeneous sample model in the collection stage, thereby improving the consistency of the water meter surface image data. Further, the image segmentation and specific image enhancement processing logic output the water meter calibration image data, which ensures that the water meter calibration system can perform calibration operation with high-precision water meter surface image, effectively optimizes the water meter surface image acquisition process, and further optimizes the precision of the calibration result of the water meter calibration system.

[0091] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A visual data acquisition system for water meter calibration, characterized in that, include: The matrix sensing module is used to sense the distance information of the water meter surface and to construct a heterogeneous model of the water meter surface based on the distance information. The modeling module is used to build heterogeneous sample models of water surface. The analysis module is used to analyze the similarity between the heterogeneous water surface model built in real time by the matrix perception module and the heterogeneous water surface sample model built by the modeling module. Based on the similarity analysis results, the module performs the acquisition operation of water surface image data. The selection module is used to traverse the water meter surface image data collected by the analysis module and select one water meter surface image data as the image data for water meter verification. The preprocessing module is used to receive water meter surface image data from the selection module as image data for water meter verification, and to perform segmentation and enhancement processing on the image data. The output module is used to acquire the enhanced water meter surface image and transmit the enhanced water meter surface image to the water meter calibration system that is pre-connected to the system. The matrix sensing module is integrated by several miniature ranging sensors, which are arranged at equal intervals in the horizontal and vertical directions, and the ranging ends of each miniature ranging sensor are all on the same plane. During the output process of the water meter from the water meter production equipment, when the water meter is delivered by the output conveyor belt to the area directly below the matrix sensing module, the matrix sensing module is triggered to operate. The matrix sensing module is connected to a robotic arm. When the matrix sensing module is triggered to run, the robotic arm controls the matrix sensing module to move horizontally in a spiral motion. During the horizontal spiral motion, the robotic arm senses the distance information of the water meter surface based on a specified frequency. The specified frequency is set to be no more than 10 times / second, and the horizontal spiral motion speed is no more than 1mm / second. The maximum diameter of the horizontal spiral motion path is no more than 3mm. The similarity analysis logic between the heterogeneous water meter surface model and the heterogeneous water meter surface sample model in the analysis module is expressed as follows: ； In the formula: denoted as , representing the similarity between heterogeneous water surface model a and heterogeneous water surface sample model b; , representing the total number of vertex control groups in heterogeneous water surface model a and heterogeneous water surface sample model b. is the offset distance of the vertex of the heterogeneous model a originating from the water surface in the i-th control group relative to the vertex of the heterogeneous sample model b originating from the water surface; Let be the longest edge of the base of the local model where the vertex of the heterogeneous model a originating from the water surface in the i-th control group is located, and let be the longest edge of the base of the local model where the vertex of the heterogeneous model b originating from the water surface in the i-th control group. In each model, each vertex is also the vertex of each drawn line segment. The vertex control group consists of two model vertices, which are respectively derived from the heterogeneous water surface model a and the heterogeneous water surface sample model b. In addition, the two vertices in a vertex control group correspond to the same miniature ranging sensor. The local model where the vertex is located is the local three-dimensional model of the heterogeneous water surface model a and the heterogeneous water surface sample model b, where the side of each vertex is a triangle.

2. The visual data acquisition system for water meter calibration process according to claim 1, characterized in that, During the operation of the matrix sensing module, each time distance information on the water meter surface is sensed, a heterogeneous model of the water meter surface is constructed simultaneously. To obtain the ranging results of each miniature ranging sensor, in three-dimensional space, draw longitudinal line segments representing each ranging result based on a plane, ensuring that the bottom ends of the drawn line segments are on the same plane, and that each bottom end is consistent with the arrangement of the miniature ranging sensors. Furthermore, the distance from the top end to the bottom end of each drawn line segment is consistent with the corresponding ranging result. Based on the adjacent interconnection of the vertices of each line segment, a closed three-dimensional figure composed of several triangular faces is obtained, which is denoted as the heterogeneous model of the water meter surface.

3. The visual data acquisition system for water meter calibration process according to claim 1, characterized in that, During the modeling module operation phase, the system user places the sample water meter on the output conveyor belt, with the sample water meter directly below the matrix sensing module. The matrix sensing module senses the distance information of the sample water meter surface and constructs a heterogeneous sample model of the water meter surface based on the distance information.

4. The visual data acquisition system for water meter calibration process according to claim 1, characterized in that, In the miniature ranging sensor of the matrix sensing module, a miniature high-definition industrial camera is deployed in the space between the miniature ranging sensor at the center position, and the high-definition industrial camera collects image data of the water surface. The miniature high-definition industrial camera operates synchronously with the matrix sensing module for each cycle, and obeys the following: When the matrix sensing module runs for the first time, the miniature high-definition industrial camera synchronously acquires an image of the water surface. The analysis module follows the matrix sensing module in each run, continuously performing similarity analysis. Starting from the second run of the matrix sensing module, the similarity analysis results are recorded. When the similarity analysis result is greater than the previous similarity analysis result, the miniature high-definition industrial camera runs once to collect water meter surface image data, and each collected water meter surface image data is marked with its corresponding similarity analysis result; If the result of each similarity analysis is less than or equal to the result of the previous similarity analysis, the data acquisition operation of the water meter surface image will not be performed; The analysis module includes a storage unit for storing water meter surface image data collected by the miniature high-definition industrial camera.

5. The visual data acquisition system for water meter calibration process according to claim 1, characterized in that, When traversing water meter image data, the selection module uses the water meter surface image data stored in the storage unit as the traversal target group, performs the traversal operation, and selects water meter surface image data based on the similarity analysis results of the labels of each water meter surface image data. The water meter surface image data with the highest similarity analysis result is selected, and the remaining water meter surface image data is deleted from the storage unit as the deletion target.

6. The visual data acquisition system for water meter calibration process according to claim 1, characterized in that, The system is pre-connected to any existing water meter verification system that verifies the appearance of a water meter through images of its surface.

7. The visual data acquisition system for water meter calibration process according to claim 1, characterized in that, The matrix perception module and modeling module are interconnected with the analysis module via a wireless network. The analysis module has a storage unit interconnected with it via a wireless network. The analysis module is interconnected with the selection module via a wireless network. The selection module is interconnected with the storage unit via a wireless network. The selection module is interconnected with the preprocessing module and the output module via a wireless network.

8. A method for visual data acquisition during water meter calibration, wherein the method is an implementation method of the visual data acquisition system for water meter calibration as described in any one of claims 1-7, characterized in that, Includes the following steps: Step 1: Construct a ranging matrix, apply the ranging matrix to sense the distance information of the water meter surface, and construct a heterogeneous model of the water meter surface based on the sensed distance information. Step 2: Construct a heterogeneous sample model of the water meter surface by combining sample water meters with a ranging matrix, and collect water meter surface image data in real time based on the similarity analysis between the heterogeneous sample model and the heterogeneous model of the water meter surface. Step 3: Mark the similarity analysis results for each collected water meter surface image data. Based on the similarity analysis results of the marking of each collected water meter surface image data, select one water meter surface image data and use the selected water meter surface image data as the image data for water meter verification. Step 4: Segment and enhance the water meter surface image data used for water meter verification to obtain the water meter surface image; Step 5: Forward the acquired water meter surface image to the preset water meter verification system, and complete the subsequent water meter appearance verification work through the water meter verification system.

Images