Visual identification method and system for water meter dial reading

Abstract

The invention discloses a water meter dial reading visual identification method and system, and relates to the field of water meter reading.The water meter dial reading visual identification system comprises a camera module used for collecting water meter dial image data and storing the water meter dial image data; the reference image cloud library is used for storing digital images in the same form as the water meter dial reading; the segmentation module is used for receiving the water meter dial image data stored in the camera module and segmenting the water meter dial image data to obtain a digital area image; according to the method, the user handheld device is used for collecting the water meter dial plate image and automatically marking the user number and time, image preprocessing, contour extraction and self-adaptive zooming are combined, a digital display area is accurately positioned, automatic and intelligent recognition of water meter reading is achieved through multi-technology fusion, and the meter reading efficiency and the data reliability are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of water meter reading, and in particular to a visual recognition method and system for water meter dial reading. Background Art

[0002] A water meter is an instrument that measures water flow and is typically installed on water pipes. Traditional water meters require manual reading, while smart water meters can transmit data remotely, facilitating measurement and management. They are widely used in homes and industries, serving as a crucial tool for water resource measurement.

[0003] The invention patent application with application number 202411920260.0 discloses a water meter dial reading recognition method, comprising: in response to a trigger reading request instruction, collecting an original water meter image; extracting a feature point set of the original water meter image, comparing the feature point set of the original water meter image with a pre-stored feature mapping table to determine a dial image segmentation rule, and segmenting the original water meter image according to the corresponding dial image segmentation rule to obtain a number of image block data; when it is determined that the boundary of each image block data is clear and the boundary of the valid information in each image block data is clear, directly identifying the valid information in each image block data to obtain the water meter dial reading; or, when it is determined that the boundary of each image block data is clear and the boundary of the valid information in at least one image block data is unclear, the image block data with unclear boundary of the valid information is segmented according to a predetermined The rule is sent to the server side so that the server side performs image recognition according to the first method to obtain the water meter dial reading; or, when it is determined that the data boundary of any image block is unclear, the original water meter image is sent to the server side so that the server side performs image recognition according to the second method to obtain the water meter dial reading and the recognition accuracy; in response to only receiving the water meter dial reading sent by the server side, the water meter dial reading is filled in the corresponding position; or, in response to receiving the water meter dial reading and the recognition accuracy sent by the server side, it is determined according to the recognition accuracy whether to automatically fill in the water meter dial reading or manually enter the water meter dial reading. This application aims to solve the problem that "in some scenarios where real-time data acquisition is required, for meter readers, this method of relying on cloud server resources has low real-time performance and reduces work efficiency."

[0004] However, although existing technologies can replace manual meter reading to a certain extent, manual meter reading is still required in special circumstances, such as water meter network failure. However, manual meter reading cannot provide fast readings due to factors such as dirt and poor lighting at the location of the water meter. Therefore, a visual recognition method and system for water meter dial readings are proposed. Summary of the Invention

[0005] In view of the above-mentioned shortcomings of the prior art, the present invention provides a method and system for visual recognition of water meter dial readings, which can effectively solve the problems of the prior art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: The present invention discloses a visual recognition system for water meter dial readings, comprising: The camera module is used to collect water meter dial image data and store the water meter dial image data; the reference image cloud library is used to store digital images with the same form as the water meter dial reading; the segmentation module is used to receive the water meter dial image data stored in the camera module, and segment the water meter dial image data to obtain digital area images; the comparison module is used to obtain the digital area images segmented and obtained in the segmentation module, and use the digital area images as comparison targets to perform similarity comparison with the digital images stored in the reference image cloud library to obtain digital images with the highest similarity to each digital area image; the arrangement module is used to receive the digital images obtained in the comparison module, and arrange each digital image based on the position of the digital area image corresponding to each digital image in the comparison module in the water meter dial area image; the recording and verification module is used to receive the digital image arrangement results in the arrangement module, and distinguish, record and verify the digital image arrangement results.

[0007] Furthermore, during the operation phase of the camera module, the user holds the camera module in front of the water meter dial and collects image data of the water meter dial in conjunction with focus and exposure. A pre-processing unit and a storage unit are provided at the lower level of the camera module. The pre-processing unit is used to traverse the water meter surface image data and extract the water meter dial area image from the water meter dial image data. The storage unit is used to receive the water meter dial area image extracted by the pre-processing unit and store the water meter dial area image. Among them, each time the camera module collects water meter dial image data, the water meter dial image data is synchronously marked with the text mark of the source household number and the collection timestamp. When the water meter dial image data is extracted based on the water meter dial area image by the preprocessing unit and stored in the storage unit, the water meter dial area image is still marked based on the household number text mark and collection timestamp mark of the water meter dial image data that is the source of the water meter dial area image.

[0008] Furthermore, during the operation phase of the pre-processing unit, the water meter dial image data collected by the camera module is received, the water meter dial image data is converted into a grayscale image, a water meter dial contour image is extracted from the grayscale image, a maximum contour is picked up from the water meter dial contour image, and the water meter dial image data is segmented using the maximum contour at a corresponding position in the water meter dial image data. The local image within the segmented maximum contour is recorded as a water meter dial area image. Among them, when the water meter dial area image is stored through the storage unit, the storage unit synchronizes the marking information of the water meter dial area images in the table, and distinguishes and stores the water meter dial area images according to the household number text mark based on the marking information, and sorts and stores them according to the collection timestamp mark.

[0009] Furthermore, the camera module has a built-in sensing component, which includes a distance sensor and a gyroscope. The number of the distance sensors is not less than three, and the number of the gyroscope is one. The camera module is manually switched on or off by the user. When the camera module is turned on, the sensing component operates synchronously. The user holds the camera module facing the water meter dial and moves it around the camera module at a constant speed. The movement path is in the shape of "∞". The gyroscope and distance sensor in the sensing component sense their own status information in real time. When any of the following forms is sensed, the camera module is triggered to operate and collect image data of the water meter dial: Form 1: The gyroscope senses that the device is in a vertical position, and the distance sensors all sense the same thing, all within the preset sensing range. Form 2: The gyroscope perceives itself as being horizontal, and the perception results of each distance sensor are consistent and within the preset perception distance range.

[0010] Furthermore, there are nine digital images stored in the reference image cloud library, representing 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9 respectively. Each digital image is located in the center of a preset digital frame, and a water meter dial image frame is also provided around the digital frame. The segmentation module operates on the received water meter dial image data, i.e., the water meter dial region image most recently stored in the storage unit. Before segmenting the water meter dial image data to obtain the digital region image, the segmentation module synchronously performs a geometric scaling operation on the water meter dial region image, i.e., scaling the water meter dial region image with reference to the water meter dial image frame so that the water meter dial region image is equal in size to the water meter dial image frame and their edges overlap; Among them, the water meter dial image frame refers to the digital display area in the actual water meter dial, and several digital frames are set at the same position. The segmentation module segments the water meter dial image data to obtain the digital area image, that is, in the water meter dial area image after the zoom operation is completed, the water meter dial area image is segmented with reference to the corresponding positions of the several digital frames set inside the water meter dial image frame to obtain several area images corresponding to the digital frames, which are recorded as digital area images.

[0011] Furthermore, before the segmentation module executes the segmentation to obtain the digital area image, the received water meter dial area image is simultaneously subjected to definition enhancement processing: ; Where: The water meter dial area image output for clarity enhancement processing; This is the water meter dial area image after adaptive median filtering denoising; is the sharpening strength factor; is the local contrast enhancement factor; is the Laplacian operator of the denoised image; in, , represents the local standard deviation at the point (x, y), Represents the adjustment factor, which is used to control the degree of contrast enhancement. represents the local mean at the point (x, y), Represents a very small constant.

[0012] Furthermore, in the comparison module, the similarity comparison logic between the digital region image and the digital image is expressed as: ; Where: is the similarity between the digital region image and the digital image; 、 is the weight; is the Hausdorff distance of the scale skeleton point set of the digital image / digital region image, that is, the scale skeleton matching degree; are the maximum and minimum values of Hausdorff distance obtained through digital image statistics; is the total number of sector areas; is the normalized angle distribution vector of the i-th sector area of the digital image; is the normalized angle distribution vector of the i-th sector area of the digital area image; Here, a similarity comparison is performed on each combination of the digital region image and the digital image based on the above formula.

[0013] Furthermore, the recording and verification module records the digital image arrangement results, that is, records the arranged digital images, and each group of recorded digital images carries the same tag of the water meter dial area image from which they originated. When recording the digital image arrangement results, the records are distinguished based on the household number in the tag content, and the records are arranged based on the acquisition timestamp in the tag content; The verification target of the recording and verification module for the digital image arrangement result is: for each distinguished recording interval, the verification operation is: reading each group of digital images arranged and recorded in the recording interval, and sorting each reading result based on the arrangement position of each group of corresponding digital images. When the sorted reading results show a continuous upward trend or no growth trend, the verification result is correct; otherwise, the verification result is incorrect.

[0014] Furthermore, the lower level of the camera module is interactively connected with the preprocessing unit and the storage unit through a wireless network, the camera module is interactively connected with the reference image cloud library and the segmentation module through a wireless network, the segmentation module is interactively connected with the storage unit through a wireless network, the segmentation module is interactively connected with the comparison module and the arrangement module through a wireless network, and the arrangement module is interactively connected with the recording and verification module through a wireless network.

[0015] On the other hand, a method for visually recognizing a water meter dial reading includes: Collect water meter dial image data and store the water meter dial image data; create a reference image cloud library and use the reference image cloud library to store digital images; obtain the stored water meter dial image data and segment the water meter dial image data to obtain a digital area image; use the digital area image to compare with the digital images stored in the reference image cloud library to obtain the digital image with the highest similarity to the digital area image; obtain the water meter dial reading by combining the obtained digital images, and record and verify the reading.

[0016] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: The present invention provides a method and system for visual recognition of water meter dial readings. During execution, the method and system collect water meter dial images through a user's handheld device and automatically mark the household number and time. Combined with image preprocessing, contour extraction and adaptive scaling, the digital display area is accurately located. Based on a multi-sensor posture perception mechanism, shooting can be automatically triggered when the device is in a horizontal or vertical stable posture to ensure image acquisition quality. Through the built-in digital image cloud library and the similarity algorithm that integrates scale skeleton matching and angular distribution characteristics, efficient comparison and recognition of digital area images are achieved, and readings are arranged according to image positions. The system can also verify the continuity of reading results based on time series to ensure data accuracy. The entire solution breaks through the traditional manual meter reading mode, realizes automatic and intelligent recognition of water meter readings through the integration of multiple technologies, and effectively improves meter reading efficiency and data reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0018] Figure 1 The following is a schematic diagram of the structure of a visual recognition system for reading water meter dials; Figure 2The figure is a flow chart of a method for visual recognition of water meter dial readings. DETAILED DESCRIPTION

[0019] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] The present invention will be further described below with reference to the embodiments.

[0021] Example 1: A visual recognition system for water meter dial readings in this embodiment, such as Figure 1 Shown, including: A camera module is used to collect and store water meter dial image data; During the camera module operation phase, the user holds the camera module in front of the water meter dial and collects image data of the water meter dial in conjunction with focus and exposure. The camera module is provided with a pre-processing unit and a storage unit at the lower level. The pre-processing unit is used to traverse the water meter surface image data and extract the water meter dial area image from the water meter dial image data. The storage unit is used to receive the water meter dial area image extracted by the pre-processing unit and store the water meter dial area image. Each time the camera module collects water meter dial image data, the water meter dial image data is synchronously marked with a text mark of the source household number and a collection timestamp. When the water meter dial image data is extracted based on the water meter dial area image by the preprocessing unit and stored in the storage unit, the water meter dial area image is still marked based on the household number text mark and the collection timestamp of the water meter dial image data of the source water meter dial area image. During the operation phase of the preprocessing unit, the water meter dial image data collected by the camera module is received, the water meter dial image data is converted into a grayscale image, a water meter dial contour image is extracted from the grayscale image, a maximum contour is picked up from the water meter dial contour image, and the water meter dial image data is segmented using the maximum contour at the corresponding position in the water meter dial image data. The local image within the segmented maximum contour is recorded as the water meter dial area image; When the water meter dial area image is stored in the storage unit, the storage unit synchronizes the marking information of each water meter dial area image in the table, distinguishes and stores the water meter dial area image according to the household number text mark based on the marking information, and sorts and stores them according to the acquisition time stamp mark; The camera module has a built-in sensing component, which includes a distance sensor and a gyroscope. The number of distance sensors set is no less than three, and the number of gyroscopes set is one. The camera module is manually switched on or off by the user. When the camera module is turned on, the sensing component runs synchronously. The user holds the camera module facing the water meter dial and moves it around the water meter at a constant speed. The movement path forms an "∞" shape. The gyroscope and distance sensor in the sensing component sense their own status information in real time. When any of the following conditions is detected, the camera module is triggered to collect image data of the water meter dial: Form 1: The gyroscope senses that the device is in a vertical position, and the distance sensors all sense the same thing, all within the preset sensing range. Form 2: The gyroscope senses that it is horizontal, and the sensing results of each distance sensor are consistent and within the preset sensing distance range; Reference image cloud library, used to store digital images with the same form as the water meter dial readings; There are nine digital images stored in the reference image cloud library, representing 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9 respectively. Each digital image is located in the center of a preset digital frame, and a water meter dial image frame is also set outside the digital frame; The segmentation module operates on the received water meter dial image data, i.e., the water meter dial region image most recently stored in the storage unit. Before segmenting the water meter dial image data to obtain the digital region image, the segmentation module synchronously performs a geometric scaling operation on the water meter dial region image, i.e., scaling the water meter dial region image with reference to the water meter dial image frame so that the water meter dial region image is equal in size to the water meter dial image frame and their edges overlap. The water meter dial image frame has a reference to the digital display area of the actual water meter dial, and a plurality of digital frames are set at the same position. The segmentation module segments the water meter dial image data to obtain a digital area image, that is, after the zoom operation is completed, the water meter dial area image is segmented with reference to the corresponding positions of the plurality of digital frames set inside the water meter dial image frame, so as to obtain a plurality of area images corresponding to the digital frames, which are recorded as digital area images. A segmentation module is used to receive the water meter dial image data stored in the camera module and segment the water meter dial image data to obtain a digital area image; Before the segmentation module performs segmentation to obtain the digital area image, the received water meter dial area image is simultaneously enhanced: ; Where: The water meter dial area image output for clarity enhancement processing; This is the water meter dial area image after adaptive median filtering denoising; is the sharpening strength factor; is the local contrast enhancement factor; is the Laplacian operator of the denoised image; in, , represents the local standard deviation at the point (x, y), Represents the adjustment factor, which is used to control the degree of contrast enhancement. represents the local mean at the point (x, y), represents a very small constant; The above logic formula is used to enhance the clarity of the water meter dial area image to improve the image quality and thus improve the accuracy of the system operation output results; A comparison module is used to obtain the digital region image obtained by segmentation in the segmentation module, and use the digital region image as a comparison target to perform similarity comparison with the digital images stored in the reference image cloud library to obtain the digital image with the highest similarity to each digital region image; In the comparison module, the similarity comparison logic between the digital region image and the digital image is expressed as: ; Where: is the similarity between the digital region image and the digital image; 、 is the weight; is the Hausdorff distance of the scale skeleton point set of the digital image / digital region image, that is, the scale skeleton matching degree; are the maximum and minimum values of Hausdorff distance obtained through digital image statistics; is the total number of sector areas; is the normalized angle distribution vector of the i-th sector area of the digital image; is the normalized angle distribution vector of the i-th sector area of the digital area image; Wherein, based on the above formula, a similarity comparison is performed on each combination of the digital region image and the digital image; By using the similarity calculation formula above, the most similar digital image is matched for each digital area image, thereby finally achieving water meter reading; an arrangement module for receiving the digital images obtained by the comparison module and arranging the digital images based on the position of the digital area images corresponding to the digital images in the comparison module and the water meter dial area image; The recording and verification module is used to receive the digital image arrangement results from the arrangement module, distinguish, record and verify the digital image arrangement results; The recording and verification module records the digital image arrangement results, that is, the arrangement of digital images. Each set of recorded digital images carries the same tag as the source water meter dial area image. When recording the digital image arrangement results, the records are distinguished based on the household number in the tag content and arranged based on the acquisition timestamp in the tag content. The verification goal of the recording and verification module for the digital image arrangement result is: for each recording interval, the verification operation is: reading each group of digital images arranged and recorded in the recording interval, sorting each reading result based on the arrangement position of each group of corresponding digital images, and if the sorted reading results show a continuous upward trend or no upward trend, the verification result is correct, otherwise, the verification result is wrong; The lower level of the camera module is interactively connected with the preprocessing unit and the storage unit through a wireless network, the camera module is interactively connected with the reference image cloud library and the segmentation module through a wireless network, the segmentation module is interactively connected with the storage unit through a wireless network, the segmentation module is interactively connected with the comparison module and the arrangement module through a wireless network, and the arrangement module is interactively connected with the recording and verification module through a wireless network.

[0022] In this embodiment, the camera module operates to collect water meter dial image data and stores the water meter dial image data. The pre-processing unit synchronously traverses the water meter surface image data and extracts the water meter dial area image from the water meter dial image data. The storage unit receives the water meter dial area image extracted by the pre-processing unit in real time and stores the water meter dial area image. The reference image cloud library is post-operated to store digital images of the same form as the water meter dial reading. The segmentation module then receives the water meter dial image data stored in the camera module, segments the water meter dial image data to obtain the digital area image, and compares the water meter dial image data. The module further obtains the digital area image obtained by segmentation in the segmentation module, and uses the digital area image as a comparison target to perform similarity comparison with the digital images stored in the reference image cloud library to obtain the digital image with the highest similarity to each digital area image. The module receives the digital image obtained in the comparison module through the arrangement module, and arranges each digital image based on the position of the digital area image corresponding to each digital image in the comparison module in the water meter dial area image. Finally, the arrangement result of the digital image in the arrangement module is received through the recording and verification module, and the arrangement result of the digital image is distinguished, recorded and verified.

[0023] Through the operation of the above system, manual meter reading is assisted, and the meter reading work is simplified to the work of water meter image collection, which effectively improves the efficiency and accuracy of meter reading.

[0024] Example 2: In terms of specific implementation, based on Example 1, this example refers to Figure 2The visual recognition system for water meter dial reading in Example 1 is further described in detail: A method for visually recognizing water meter dial readings, comprising: Step 1: Collect water meter dial image data and store the water meter dial image data; Step 2: Create a reference image cloud library and use the reference image cloud library to store digital images; Step 3: Obtain the stored water meter dial image data, and segment the water meter dial image data to obtain a digital area image; Step 4: Compare the digital region image with the digital images stored in the reference image cloud library to obtain the digital image with the highest similarity to the digital region image; Step 5: Obtain the water meter dial reading by combining the acquired digital images, and record and verify the reading.

[0025] In summary, during the execution of the method and system in the above embodiments, the user's handheld device is used to collect the water meter dial image and automatically mark the household number and time. Combined with image preprocessing, contour extraction and adaptive scaling, the digital display area is accurately located. Based on the multi-sensor posture perception mechanism, the shooting can be automatically triggered when the device is in a horizontal or vertical stable posture to ensure the quality of image acquisition. Through the built-in digital image cloud library and the similarity algorithm that integrates the scale skeleton matching degree and the angle distribution characteristics, efficient comparison and recognition of digital area images are achieved, and the reading arrangement is completed according to the image position. The system can also verify the continuity of the reading results based on the time series to ensure data accuracy. The whole solution breaks through the traditional manual meter reading mode and realizes the automatic and intelligent recognition of water meter readings through the integration of multiple technologies, effectively improving meter reading efficiency and data reliability.

[0026] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A visual recognition system for water meter dial readings, characterized in that: include: A camera module is used to collect and store water meter dial image data; Reference image cloud library, used to store digital images with the same form as the water meter dial readings; A segmentation module is used to receive the water meter dial image data stored in the camera module and segment the water meter dial image data to obtain a digital area image; A comparison module is used to obtain the digital region image obtained by segmentation in the segmentation module, and use the digital region image as a comparison target to perform similarity comparison with the digital images stored in the reference image cloud library to obtain the digital image with the highest similarity to each digital region image; an arrangement module for receiving the digital images obtained by the comparison module and arranging the digital images based on the position of the digital area images corresponding to the digital images in the comparison module and the water meter dial area image; The recording and verification module is used to receive the digital image arrangement results from the arrangement module, distinguish, record and verify the digital image arrangement results.

2. A visual recognition system for water meter dial readings according to claim 1, characterized in that: During the operation phase of the camera module, the user holds the camera module in front of the water meter dial and collects image data of the water meter dial in conjunction with focus and exposure. A preprocessing unit and a storage unit are provided at the lower level of the camera module. The preprocessing unit is used to traverse the water meter surface image data and extract the water meter dial area image from the water meter dial image data. The storage unit is used to receive the water meter dial area image extracted by the preprocessing unit and store the water meter dial area image. Among them, each time the camera module collects water meter dial image data, the water meter dial image data is synchronously marked with the text mark of the source household number and the collection timestamp. When the water meter dial image data is extracted based on the water meter dial area image by the preprocessing unit and stored in the storage unit, the water meter dial area image is still marked based on the household number text mark and collection timestamp mark of the water meter dial image data that is the source of the water meter dial area image.

3. A visual recognition system for water meter dial readings according to claim 2, characterized in that: During the operation phase of the preprocessing unit, the water meter dial image data collected by the camera module is received, the water meter dial image data is converted into a grayscale image, a water meter dial contour image is extracted from the grayscale image, a maximum contour is picked up from the water meter dial contour image, and the water meter dial image data is segmented using the maximum contour at a corresponding position in the water meter dial image data. The local image within the segmented maximum contour is recorded as a water meter dial area image; Among them, when the water meter dial area image is stored through the storage unit, the storage unit synchronizes the marking information of the water meter dial area images in the table, and distinguishes and stores the water meter dial area images according to the household number text mark based on the marking information, and sorts and stores them according to the collection timestamp mark.

4. A visual recognition system for water meter dial readings according to claim 1, characterized in that: The camera module has a built-in sensing component, which includes a distance sensor and a gyroscope. The number of the distance sensors is not less than three, and the number of the gyroscope is one. The operating state of the camera module is manually switched on or off by the user end; When the camera module is turned on, the sensing component operates synchronously. The user holds the camera module facing the water meter dial and moves it around the camera module at a constant speed. The movement path is in the shape of "∞". The gyroscope and distance sensor in the sensing component sense their own status information in real time. When any of the following forms is sensed, the camera module is triggered to operate and collect image data of the water meter dial: Form 1: The gyroscope senses that the device is in a vertical position, and the distance sensors all sense the same thing, all within the preset sensing range. Form 2: The gyroscope perceives itself as being horizontal, and the perception results of each distance sensor are consistent and within the preset perception distance range.

5. A visual recognition system for water meter dial readings according to claim 1, characterized in that: There are nine digital images stored in the reference image cloud library, representing 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9 respectively. Each digital image is located in the center of a preset digital frame, and a water meter dial image frame is also provided outside the digital frame; The segmentation module operates on the received water meter dial image data, i.e., the water meter dial region image most recently stored in the storage unit. Before segmenting the water meter dial image data to obtain the digital region image, the segmentation module synchronously performs a geometric scaling operation on the water meter dial region image, i.e., scaling the water meter dial region image with reference to the water meter dial image frame so that the water meter dial region image is equal in size to the water meter dial image frame and their edges overlap; Among them, the water meter dial image frame refers to the digital display area in the actual water meter dial, and several digital frames are set at the same position. The segmentation module segments the water meter dial image data to obtain the digital area image, that is, in the water meter dial area image after the zoom operation is completed, the water meter dial area image is segmented with reference to the corresponding positions of the several digital frames set inside the water meter dial image frame to obtain several area images corresponding to the digital frames, which are recorded as digital area images.

6. A visual recognition system for water meter dial readings according to claim 5, characterized in that: Before the segmentation module executes segmentation to obtain the digital area image, it simultaneously performs definition enhancement processing on the received water meter dial area image: ； Where: The water meter dial area image output for clarity enhancement processing; This is the water meter dial area image after adaptive median filtering denoising; is the sharpening strength factor; is the local contrast enhancement factor; is the Laplacian operator of the denoised image; in, , represents the local standard deviation at the point (x, y), Represents the adjustment factor, which is used to control the degree of contrast enhancement. represents the local mean at the point (x, y), Represents a very small constant.

7. A visual recognition system for water meter dial readings according to claim 1, characterized in that: In the comparison module, the similarity comparison logic between the digital region image and the digital image is expressed as: ； Where: is the similarity between the digital region image and the digital image; 、 is the weight; is the Hausdorff distance of the scale skeleton point set of the digital image / digital region image, that is, the scale skeleton matching degree; are the maximum and minimum values of Hausdorff distance obtained through digital image statistics; is the total number of sector areas; is the normalized angle distribution vector of the i-th sector area of the digital image; is the normalized angle distribution vector of the i-th sector area of the digital area image; Here, a similarity comparison is performed on each combination of the digital region image and the digital image based on the above formula.

8. A visual recognition system for water meter dial readings according to claim 1, characterized in that: The recording and verification module records the digital image arrangement results, that is, records the arranged digital images, each set of recorded digital images carries the same mark as the source water meter dial area image, and when recording the digital image arrangement results, the records are distinguished based on the household number in the mark content, and the records are arranged based on the acquisition timestamp in the mark content; The verification target of the recording and verification module for the digital image arrangement result is: for each distinguished recording interval, the verification operation is: reading each group of digital images arranged and recorded in the recording interval, and sorting each reading result based on the arrangement position of each group of corresponding digital images. When the sorted reading results show a continuous upward trend or no growth trend, the verification result is correct; otherwise, the verification result is incorrect.

9. A visual recognition system for water meter dial readings according to claim 1, characterized in that: The lower level of the camera module is interactively connected with the preprocessing unit and the storage unit through a wireless network, the camera module is interactively connected with the reference image cloud library and the segmentation module through a wireless network, the segmentation module is interactively connected with the storage unit through a wireless network, the segmentation module is interactively connected with the comparison module and the arrangement module through a wireless network, and the arrangement module is interactively connected with the recording and verification module through a wireless network.

10. A method for visually recognizing water meter dial readings, the method being an implementation of a system for visually recognizing water meter dial readings as claimed in any one of claims 1 to 9, characterized in that: include: Step 1: Collect water meter dial image data and store the water meter dial image data; Step 2: Create a reference image cloud library and use the reference image cloud library to store digital images; Step 3: Obtain the stored water meter dial image data, and segment the water meter dial image data to obtain a digital area image; Step 4: Compare the digital region image with the digital images stored in the reference image cloud library to obtain the digital image with the highest similarity to the digital region image; Step 5: Obtain the water meter dial reading by combining the acquired digital images, and record and verify the reading.

Images