Intelligent meter reading method and system based on AI vision and electronic equipment

Abstract

The invention relates to an intelligent meter reading method and system based on AI vision and electronic equipment. The intelligent meter reading system comprises a water meter, a camera device, a server and a water meter cleaning device, the method is applied to a server and comprises the following steps: acquiring a water meter image shot by a camera device in real time, analyzing the water meter image and determining whether the water meter has stains; if it is determined that the stains exist, the water meter type of the water meter is determined according to the water meter image; determining a data distribution condition of the water meter according to the water meter type; according to the data distribution condition, whether the stains affect data reading or not is determined; if the data reading is influenced, the water meter cleaning device is started to clean the water meter to read the value of the water meter so as to ensure the normal work of the water meter and the accurate reading of the data. Automatic and intelligent water meter monitoring and cleaning are achieved, the working efficiency is improved, and manual intervention and operation errors are reduced. Stains can be found and treated in time, and better service is provided for water supply companies and users.

Description

Technical Field

[0001] The present application relates to the technical field of water meter reading, and in particular to an AI vision-based intelligent meter reading method, system and electronic equipment. Background Art

[0002] Water meters can measure the flow and cumulative flow of water, thus helping users and water supply companies understand water usage and calculate water charges and water consumption. Currently, manual door-to-door water meter reading is the most common method. Since water meters are set up in too scattered locations and in large numbers, a large number of people are required to collect data on site.

[0003] This seriously affects the efficiency of meter reading. In addition, water meter readings will fluctuate greatly due to factors such as changes in water consumption and pipeline leakage. Therefore, manual meter reading will also have errors at a high frequency.

[0004] To this end, a method has been designed that can obtain water meter images by photographing the water meter with a camera and then analyze them to achieve meter reading. However, this method can only achieve accurate readings when the water meter dial is clear. When there are stains, the data may be blocked, and the data on the water meter cannot be read correctly, resulting in meter reading errors. In this case, manual cleaning of stains is still required before re-reading, and manual intervention is still required, and manual intervention cannot be eliminated. Summary of the invention

[0005] The present application provides an AI vision-based intelligent meter reading method, system and electronic device, which have solved the above-mentioned problem of low meter reading accuracy.

[0006] In a first aspect, the present application provides an AI vision-based intelligent meter reading method, which is applied to an intelligent meter reading device, wherein the intelligent meter reading device includes a water meter, a camera device, a server, and a water meter cleaning device; the method is applied to the server, and the method includes:

[0007] Acquire a water meter image captured in real time by the camera device, analyze the water meter image, and determine whether there is stain on the water meter;

[0008] If it is determined that there is stain, determining the water meter type of the water meter according to the water meter image;

[0009] Determining data distribution of the water meter according to the type of the water meter;

[0010] Determining whether the stain affects data reading according to the data distribution;

[0011] If data reading is affected, the water meter cleaning device is started to clean the water meter to read the water meter value.

[0012] Through this solution, by using a camera to capture the water meter image in real time, it is possible to promptly detect whether there are stains on the water meter and whether the stains affect the normal operation of the water meter. Once it is determined that the stains affect the data reading, the water meter cleaning device can be automatically started to clean the water meter to ensure the normal operation of the water meter and accurate data reading. This solution can realize automated and intelligent water meter monitoring and cleaning, improve work efficiency, and reduce manual intervention and operational errors. Timely detection and treatment of stains can ensure the accuracy and reliability of water meter readings and provide better services for water supply companies and users.

[0013] Optionally, determining the water meter type of the water meter according to the water meter image includes:

[0014] Recognize the water meter image and determine the outline of the water meter;

[0015] According to the profile, water meter data in a water meter information database is screened to determine a matching water meter;

[0016] Extracting features from the water meter image to obtain water meter features, and comparing the water meter features with features of the matching water meter;

[0017] According to the comparison result, determining whether there is a target water meter that matches all the water meter characteristics;

[0018] If it exists, the water meter information of the target water meter is retrieved to determine the water meter type of the water meter.

[0019] Through this solution, the extracted water meter profile information can be matched with the data in the water meter information database to quickly screen out possible matching water meters. By extracting features from water meter images, water meter features can be obtained, and these features can be compared with the features of matching water meters to further determine whether there is a target water meter that completely matches the water meter features. This helps to ensure the accuracy and reliability of data, improve work efficiency, and reduce manual intervention and operational errors.

[0020] Optionally, starting the water meter cleaning device to clean the water meter further includes:

[0021] Obtain a detail image corresponding to the stain at the data distribution position;

[0022] Binarizing the detail image to obtain a clear image, and extracting stain features of the clear image;

[0023] Clustering the stain features, analyzing the features of each cluster after clustering, and determining the characteristics of the stain;

[0024] Based on the characteristics, the type of the stain is determined.

[0025] Through this solution, detailed images, binarization processing, feature extraction, feature clustering and characteristic analysis are obtained. This step-by-step processing method can improve the accuracy and efficiency of processing. It reduces manual intervention and operational errors. By identifying the type of stains, cleaning and maintenance work can be carried out effectively, reducing unnecessary waste and pollution.

[0026] Optionally, starting the water meter cleaning device to clean the water meter includes:

[0027] Determine the best cleaning agent based on the type of stain in question;

[0028] Acquire the placement data of the water meter cleaning device; the placement data includes the placement quantity;

[0029] Determining whether the optimal cleaning agent exists in the water meter cleaning device according to the placement data;

[0030] If so, determining the optimal amount of cleaning agent to be used according to the size of the stain;

[0031] Comparing the used amount with the placed amount, and determining whether the stain can be completely removed according to the comparison result;

[0032] If it is determined that the stain can be completely removed, the water meter cleaning device is started to treat the stain.

[0033] Through this solution, the most suitable cleaning agent is selected according to the type of stain, which can improve the cleaning effect, reduce damage to the water meter, and extend the service life of the water meter. It improves cleaning efficiency, reduces manual intervention and operating errors, and also avoids damage to the water meter caused by improper manual cleaning. The optimal amount of cleaning agent is determined according to the size of the stain, and whether the stain can be completely removed is evaluated, which avoids insufficient or excessive cleaning, avoids the waste of cleaning agent, and also ensures the cleaning effect.

[0034] Optionally, the method further includes:

[0035] If it is determined that the stain cannot be completely removed, determining the cleaning degree of the current amount of items placed according to the cleaning data of the optimal cleaning agent;

[0036] Determining data exposure according to the cleaning degree and the data distribution;

[0037] Determining whether the stain affects meter reading according to the data revealed;

[0038] If there is no impact, start the water meter cleaning device to treat the stain.

[0039] Through this solution, the cleaning degree of the current amount of material placed can be accurately determined through the cleaning data of the optimal cleaning agent, and the data exposure situation can be determined based on the cleaning degree and data distribution, so as to judge whether the use of the amount of material placed for cleaning will affect the meter reading, thereby reducing the waste of the amount of material placed. At the same time, it can also improve the processing efficiency, avoid the situation where the meter reading is affected by not cleaning when the amount of material placed is insufficient, improve the automation and intelligence of the processing, improve the processing efficiency and reduce manual intervention.

[0040] Optionally, the method further includes:

[0041] Using the water meter images taken continuously during a preset period of time, determine whether there is currently any abnormal water discharge;

[0042] If there is water outflow abnormality at the current moment, obtain the precise location of the water meter at the current moment;

[0043] According to the precise positioning, the number of water outlets is determined, and sensors corresponding to the water outlets are activated to sense the water flow;

[0044] Determining whether there is an abnormal water outlet according to the water flow rate;

[0045] If it does not exist, it is determined that the pipeline is abnormal.

[0046] Through this solution, by continuously taking pictures of water meters for real-time monitoring, abnormal water discharge can be discovered in time to avoid waste and loss of water resources. If abnormal water discharge is found, the solution can obtain the precise location of the water meter, accurate to the specific water outlet, and improve the accuracy of abnormal location. By using image processing and recognition technology, abnormal water discharge can be automatically detected, avoiding the tediousness and errors of manual inspections and improving detection efficiency and accuracy.

[0047] Optionally, determining whether there is an abnormal water outlet according to the water flow rate includes:

[0048] Obtaining historical water flow data of the water meter;

[0049] determining water use habits based on the historical water flow data;

[0050] Determining the expected water flow at the current moment based on the water use habits;

[0051] Comparing the water flow rate with the expected water flow rate to determine whether there is an abnormality in the water flow rate;

[0052] If the water flow rate is abnormal, it is determined that there is an abnormal water outlet.

[0053] Through this solution, historical water flow data is analyzed to understand the user's water usage habits, so that the expected water flow at the current moment can be predicted more accurately. Comparing the actual water flow with the expected water flow can more accurately determine whether there is an abnormality in the water flow, thereby improving the accuracy of anomaly detection.

[0054] In a second aspect, the present application provides an AI vision-based intelligent meter reading system, comprising:

[0055] A stain analysis module, used to obtain the water meter image taken by the camera device in real time, analyze the water meter image, and determine whether there is stain on the water meter;

[0056] a type determination module, configured to determine the water meter type of the water meter according to the water meter image if it is determined that there is stain;

[0057] A data distribution analysis module, used to determine the data distribution of the water meter according to the type of the water meter;

[0058] A reading judgment module, used to determine whether the stain affects data reading according to the data distribution;

[0059] The water meter cleaning control module is used to start the water meter cleaning device to clean the water meter to read the water meter value if the data reading is affected.

[0060] Optionally, the type determination module is specifically used to:

[0061] Recognize the water meter image and determine the outline of the water meter;

[0062] According to the profile, water meter data in a water meter information database is screened to determine a matching water meter;

[0063] Extracting features from the water meter image to obtain water meter features, and comparing the water meter features with features of the matching water meter;

[0064] According to the comparison result, determining whether there is a target water meter that matches all the water meter characteristics;

[0065] If it exists, the water meter information of the target water meter is retrieved to determine the water meter type of the water meter.

[0066] Optionally, the intelligent meter reading system further includes a stain type analysis module, which is used to:

[0067] Obtain a detail image corresponding to the stain at the data distribution position;

[0068] Binarizing the detail image to obtain a clear image, and extracting stain features of the clear image;

[0069] Clustering the stain features, analyzing the features of each cluster after clustering, and determining the characteristics of the stain;

[0070] Based on the characteristics, the type of the stain is determined.

[0071] Optionally, the water meter cleaning control module is specifically used for:

[0072] Determine the best cleaning agent based on the type of stain in question;

[0073] Acquire the placement data of the water meter cleaning device; the placement data includes the placement quantity;

[0074] Determining whether the optimal cleaning agent exists in the water meter cleaning device according to the placement data;

[0075] If so, determining the optimal amount of cleaning agent to be used according to the size of the stain;

[0076] Comparing the used amount with the placed amount, and determining whether the stain can be completely removed according to the comparison result;

[0077] If it is determined that the stain can be completely removed, the water meter cleaning device is started to treat the stain.

[0078] Optionally, the intelligent meter reading system further includes a storage quantity analysis module, which is used to:

[0079] If it is determined that the stain cannot be completely removed, determining the cleaning degree of the current amount of items placed according to the cleaning data of the optimal cleaning agent;

[0080] Determining data exposure according to the cleaning degree and the data distribution;

[0081] Determining whether the stain affects meter reading according to the data revealed;

[0082] If there is no impact, start the water meter cleaning device to treat the stain.

[0083] Optionally, the intelligent meter reading system further includes an abnormality analysis module, which is used to:

[0084] Using the water meter images taken continuously during a preset period of time, determine whether there is currently any abnormal water discharge;

[0085] If there is water outflow abnormality at the current moment, obtain the precise location of the water meter at the current moment;

[0086] According to the precise positioning, the number of water outlets is determined, and sensors corresponding to the water outlets are activated to sense the water flow;

[0087] Determining whether there is an abnormal water outlet according to the water flow rate;

[0088] If it does not exist, it is determined that the pipeline is abnormal.

[0089] Optionally, the abnormality analysis module is specifically used for:

[0090] Obtaining historical water flow data of the water meter;

[0091] determining water use habits based on the historical water flow data;

[0092] Determining the expected water flow at the current moment based on the water use habits;

[0093] Comparing the water flow rate with the expected water flow rate to determine whether there is an abnormality in the water flow rate;

[0094] If the water flow rate is abnormal, it is determined that there is an abnormal water outlet.

[0095] In a third aspect, the present application provides an electronic device, comprising: a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the method of the first aspect.

[0096] In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program that can be loaded by a processor and execute the method of the first aspect.

[0097] The above technical solution adopted in this application has the following advantages: by taking the water meter image in real time through the camera device, it is possible to timely find out whether there are stains on the water meter and whether the stains affect the normal operation of the water meter. Once it is determined that the stains affect the data reading, the water meter cleaning device can be automatically started to clean the water meter to ensure the normal operation of the water meter and the accurate reading of the data. The scheme can realize automatic and intelligent water meter monitoring and cleaning, improve work efficiency, and reduce manual intervention and operation errors. Timely discovery and treatment of stains can ensure the accuracy and reliability of water meter readings and provide better services for water supply companies and users. According to the extracted water meter profile information, it can be matched with the data in the water meter information library to quickly screen out possible matching water meters. By extracting features from the water meter image, water meter features can be obtained, and these features are compared with the features of the matching water meter to further determine whether there is a target water meter that completely matches the water meter features, which helps to ensure the accuracy and reliability of the data, improve work efficiency, and reduce manual intervention and operation errors. Obtaining detail images, binarization processing, feature extraction, feature clustering and feature analysis, etc., this step-by-step processing method can improve the accuracy and efficiency of processing. Reduced manual intervention and operational errors. By identifying the type of stains, cleaning and maintenance work can be carried out effectively, reducing unnecessary waste and pollution. Selecting the most suitable cleaning agent according to the type of stain can improve the cleaning effect, reduce damage to the water meter, and extend the service life of the water meter. Improving cleaning efficiency, reducing manual intervention and operational errors, and also avoiding damage to the water meter caused by improper manual cleaning. Determine the amount of the best cleaning agent to be used according to the size of the stain, evaluate whether the stain can be completely removed, avoid insufficient or excessive cleaning, avoid the waste of cleaning agents, and ensure the cleaning effect. Through the cleaning data of the best cleaning agent, the cleaning degree of the current amount of material can be accurately determined, and the data exposure can be determined according to the cleaning degree and data distribution, so as to determine whether the use of the amount of material for cleaning will affect the meter reading, reduce the waste of the amount of material, and also improve the processing efficiency, avoid the situation that the meter reading is affected by not cleaning when the amount of material is insufficient, improve the automation and intelligence of the processing, improve the processing efficiency and reduce manual intervention. By continuously shooting water meter images for real-time monitoring, abnormal water discharge can be discovered in time to avoid waste and loss of water resources. If water outflow anomalies are found, the solution can obtain the precise location of the water meter, accurate to the specific water outlet, and improve the accuracy of abnormal location. By using image processing and recognition technology, water outflow anomalies can be automatically detected, avoiding the tediousness and errors of manual inspections, and improving detection efficiency and accuracy. By analyzing historical water flow data and understanding users' water use habits, it is possible to more accurately predict the expected water flow at the current moment. By comparing the actual water flow with the expected water flow, it is possible to more accurately determine whether there is an abnormality in the water flow, improving the accuracy of anomaly detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0099] Figure 1 A schematic diagram of an application scenario provided for an embodiment of the present application;

[0100] Figure 2 A flowchart of an AI vision-based intelligent meter reading method provided in one embodiment of the present application;

[0101] Figure 3 A schematic diagram of the structure of an AI vision-based intelligent meter reading device provided in one embodiment of the present application;

[0102] Figure 4 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0103] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0104] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0105] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0106] Currently, manual door-to-door water meter reading is the most common method. Since water meters are located too scattered and in large numbers, a large number of people are required to collect water meter data on site.

[0107] This seriously affects the efficiency of meter reading. In addition, water meter readings will fluctuate greatly due to factors such as changes in water consumption and pipeline leakage. Therefore, manual meter reading will also have errors at a high frequency.

[0108] To this end, a method has been designed that can obtain water meter images by photographing the water meter with a camera and then analyze them to achieve meter reading. However, this method can only achieve accurate readings when the water meter dial is clear. When there are stains, the data may be blocked, and the data on the water meter cannot be read correctly, resulting in meter reading errors. In this case, manual cleaning of stains is still required before re-reading, and manual intervention is still required, and manual intervention cannot be eliminated.

[0109] Based on this, the present application provides an AI vision-based intelligent meter reading method, system and electronic device. By taking real-time pictures of the water meter with a camera device, it is possible to promptly detect whether there are stains on the water meter and whether the stains affect the normal operation of the water meter. Once it is determined that the stains affect the data reading, the water meter cleaning device can be automatically started to clean the water meter to ensure the normal operation of the water meter and accurate reading of the data. This solution can realize automated and intelligent water meter monitoring and cleaning, improve work efficiency, and reduce manual intervention and operational errors. Timely detection and treatment of stains can ensure the accuracy and reliability of water meter readings and provide better services for water supply companies and users.

[0110] Figure 1 A schematic diagram of an application scenario provided by the present application. When reading a water meter, the intelligent meter reading method based on AI vision provided by the present application can be adopted. The present application is applied to an intelligent meter reading device, which includes a water meter, a camera device, a server, and a water meter cleaning device, wherein the water meter, the camera device, and the water meter cleaning device are arranged at a location where water use exists, and the server can be a server for managing water meter data, or any server for data management under the water service system. When reading water meter data, first obtain a water meter image through a camera device, analyze and determine whether there are stains, and when there are stains, control the water meter cleaning device to clean it for easy reading. The water meter is treated with stains in the presence of stains to achieve the purpose of accurately reading water meter data.

[0111] For specific implementation methods, please refer to the following embodiments.

[0112] Figure 2 This is a flow chart of an AI vision-based intelligent meter reading method provided in one embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes:

[0113] S201, obtaining a water meter image captured in real time by a camera device, analyzing the water meter image, and determining whether there is stain on the water meter.

[0114] The camera device is arranged at a position where the overall image of the water meter can be clearly captured.

[0115] The analysis of the water meter image can be achieved by feature extraction. The extracted water meter features are input into a preset recognition model, and the presence of stains on the water meter is determined based on the recognition results of the preset recognition model.

[0116] Among them, the preset recognition model can use a large number of stain features as samples for training, so that it can determine whether there is a stain feature by identifying the input features.

[0117] S202: If it is determined that there is stain, determine the water meter type of the water meter according to the water meter image.

[0118] There are several different types of water meters in the prior art, and the way the data of each type of water meter is displayed may be different. Before classification, the water meters can be classified according to the data display methods of several types of water meters in the prior art. If the data display methods are consistent, they can be summarized as the same type. When classifying in this way, the appearance and size can be ignored, and only the relative position of each data display area on the water meter can be considered. The relative position involved above can be judged based on the outer edge contour of the water meter. At this time, at least one of the leftmost side of the water meter, the top of the water meter, the rightmost side of the water meter, and the bottom of the water meter can be used as a reference point.

[0119] It should be noted that the above classification method is only an example. In this embodiment, it is not limited to only the above classification method. Other classification methods or type refinements can be performed according to the specific situation of the water meter.

[0120] After determining what types of water meters exist, the water meter image is analyzed according to the above classification method to determine the relative position relationship of various data in the water meter displayed on the water meter, thereby determining the water meter type of the water meter.

[0121] S203: Determine data distribution of the water meter according to the type of the water meter.

[0122] According to the water meter classification method described in the above step S202, the data display situation of the current water meter can be determined, and the area in which the data is displayed can be determined. Then, according to the water meter characteristics extracted in the above step S201, the amount of data displayed can be determined, and then according to the display method of each data when the water meter itself is designed, the data distribution situation can be determined, which may include the data type and the data distribution location.

[0123] S204: Determine whether the stain affects data reading according to the data distribution.

[0124] According to the stain feature obtained above, the position of the stain feature on the water meter is determined, and then according to the data distribution obtained above, it is determined whether there is position overlap. If so, it can be considered that data reading is affected.

[0125] S205: If data reading is affected, start the water meter cleaning device to clean the water meter to read the water meter value.

[0126] The water meter cleaning device can be arranged near the water meter. The water meter cleaning device can be provided with a plurality of cleaning agents, which can be diluted in the water meter cleaning device. When cleaning is required, the diluted cleaning agent can be sprayed onto the dial of the water meter. In addition, in order to prevent the cleaning agent from affecting the reading, a suction cup that can absorb the cleaning agent can be arranged on the water meter cleaning device. After the cleaning agent adheres to the dial for a certain period of time, the cleaning disc can be used to wipe and absorb the cleaning agent on the dial to clean it.

[0127] Through the solution provided by the present application, by taking real-time images of the water meter through a camera device, it is possible to promptly detect whether there are stains on the water meter and whether the stains affect the normal operation of the water meter. Once it is determined that the stains affect the data reading, the water meter cleaning device can be automatically started to clean the water meter to ensure the normal operation of the water meter and accurate data reading. This solution can realize automated and intelligent water meter monitoring and cleaning, improve work efficiency, and reduce manual intervention and operational errors. Timely detection and treatment of stains can ensure the accuracy and reliability of water meter readings and provide better services for water supply companies and users.

[0128] In some embodiments, a water meter image is identified to determine the outline of the water meter; based on the outline, water meter data in a water meter information database is screened to determine a matching water meter; features are extracted from the water meter image to obtain water meter features, and the water meter features are compared with features of a matching water meter; based on the comparison results, it is determined whether there is a target water meter that matches all the water meter features; if so, the water meter information of the target water meter is retrieved to determine the water meter type of the water meter.

[0129] In addition to the water meter type distinction method described in the above embodiment, the water meters may also be distinguished by their appearance. Different types of water meters may have different appearances. Therefore, the water meters may be classified by using the outlines of existing water meters.

[0130] Therefore, after obtaining the water meter image, edge detection can be performed on the image, and after the edge position is determined, edge pixels can be extracted to determine the contour features, thereby obtaining the water meter contour.

[0131] The water meter information database contains information about several existing water meters, such as models, etc., as well as corresponding images, water meter image features, appearance, information displayed on the water meter, etc. If water meter information that is consistent with the water meter profile is matched from the water meter information database, the water meter corresponding to this water meter information can be used as the matching water meter.

[0132] After obtaining the matching water meter, the water meter features can be compared with the water meter image features corresponding to the matching water meter again, so as to determine the target water meter in the matching water meter, and this target water meter can be considered to be the same type of water meter as the water meter in the water meter image. At this time, the water meter information of the target water meter is retrieved to obtain the water meter type corresponding to the water meter image.

[0133] Through the solution provided by this embodiment, the extracted water meter profile information can be matched with the data in the water meter information database to quickly screen out possible matching water meters. By extracting features from the water meter image, water meter features can be obtained, and these features can be compared with the features of the matching water meter to further determine whether there is a target water meter that completely matches the water meter features, which helps to ensure the accuracy and reliability of data, improve work efficiency, and reduce manual intervention and operational errors.

[0134] In some embodiments, a detail image corresponding to a stain at a data distribution location is obtained; the detail image is binarized to obtain a clear image, and stain features of the clear image are extracted; the stain features are clustered, and the features of each cluster after clustering are analyzed to determine the characteristics of the stain; and the type of the stain is determined based on the characteristics.

[0135] The detail image can be considered as a close-up of the stain obtained by zooming in on the stain position through the camera device. In a specific implementation, in order to ensure that the stain is fully captured, the zoom level of the camera device can be adjusted to determine that there is at least one water meter feature at the outermost edge of the image.

[0136] After obtaining the detail image, the image is binarized, for example, a global threshold method, a local threshold method, or an adaptive threshold method may be used. After the binarization is completed, the detail image may be post-processed to obtain a clear image.

[0137] The characteristics of a stain may include the texture of the stain. Different stains have different textures. By analyzing this texture characteristic, the type of stain can be determined.

[0138] Through the solution provided in this embodiment, detailed images, binarization processing, feature extraction, feature clustering and characteristic analysis are obtained. This step-by-step processing method can improve the accuracy and efficiency of the processing. It reduces manual intervention and operation errors. By identifying the type of stains, cleaning and maintenance work can be carried out effectively, reducing unnecessary waste and pollution.

[0139] In some embodiments, the best cleaning agent is determined according to the type of stain; the placement data of the water meter cleaning device is obtained; the placement data includes the placement quantity; based on the placement data, it is determined whether the best cleaning agent is present in the water meter cleaning device; if so, the usage of the best cleaning agent is determined according to the size of the stain; the usage is compared with the placement quantity, and based on the comparison result, it is determined whether the stain can be completely removed; if it is determined that the stain can be completely removed, the water meter cleaning device is started to treat the stain.

[0140] The best cleaning agent can be considered as the one that can achieve the best cleaning effect at the fastest speed. Since the environment in which the water meter is located may be complex, there may be many different types of stains. In order to improve the cleaning efficiency, the cleaning agent can be set according to the common stains on the water meter.

[0141] In a specific implementation, when installing a water meter, the type of stains that may exist in the environment of the installation location can be determined based on the installation location of the water meter. Once determined, the cleaning agent in the corresponding water meter cleaning device can be mainly the best cleaning agent for the stain type.

[0142] The placement amount can be considered as the amount of cleaning agent in the water meter cleaning device, and the usage amount can be considered as the optimal amount of cleaning agent required to clean the stains on the above-mentioned water meter.

[0143] Through the solution provided by this embodiment, the most suitable cleaning agent is selected according to the type of stain, which can improve the cleaning effect, reduce damage to the water meter, and extend the service life of the water meter. The cleaning efficiency is improved, manual intervention and operating errors are reduced, and damage to the water meter caused by improper manual cleaning can also be avoided. The optimal amount of cleaning agent is determined according to the size of the stain, and whether the stain can be completely removed is evaluated, which avoids insufficient or excessive cleaning, can avoid the waste of cleaning agent, and can also ensure the cleaning effect.

[0144] In some embodiments, if it is determined that the stain cannot be completely removed, the cleaning degree of the current amount of material placed is determined based on the cleaning data of the optimal cleaning agent; the data exposure is determined based on the cleaning degree and data distribution; based on the data exposure, it is determined whether the stain affects the meter reading; if not, the water meter cleaning device is started to treat the stain.

[0145] Since the use of detergent may cause the storage volume to continue to decrease, if it is not replenished in time, the storage volume for the last wash may be insufficient and the stains cannot be completely cleaned.

[0146] At this time, it is necessary to predict the degree of cleaning. The prediction method can be to establish a neural network model, and use the data such as the type of cleaning agent, the type and degree of stains, cleaning conditions and cleaning effects as samples to test this neural network model, so that the model can determine the degree of cleaning of stains by different amounts of cleaning agents. The amount of material is input into the above-mentioned trained neural network model to determine the degree of cleaning, and then according to the degree of cleaning, determine the data exposure of the stains that block the data after cleaning. According to this data exposure, determine whether it affects the meter reading. For example, when the tonnage of the water meter is blocked, it is considered to affect the meter reading. If the amount of material is not enough to completely clean the stains, but it does not affect the meter reading, the water meter cleaning device can continue to be started to treat the stains.

[0147] Through the scheme provided by this embodiment, the cleaning degree of the current amount of material placed can be accurately determined through the cleaning data of the optimal cleaning agent, and the data exposure situation can be determined according to the cleaning degree and the data distribution situation, so as to judge whether the meter reading will be affected after cleaning with the amount of material placed, thereby reducing the waste of the amount of material placed, and at the same time, the processing efficiency can be improved, avoiding the situation where the meter reading is affected by not cleaning when the amount of material placed is insufficient, thereby improving the automation and intelligence of the processing, improving the processing efficiency and reducing manual intervention.

[0148] In some embodiments, water meter images taken continuously during a preset period of time are used to determine whether there is currently any abnormal water outflow; if there is currently any abnormal water outflow, the precise positioning of the water meter at the current moment is obtained; based on the precise positioning, the number of water outlets is determined, and the sensors corresponding to the water outlets are activated to sense the water flow; based on the water flow, it is determined whether there is an abnormal water outlet; if not, it is determined that the pipeline is abnormal.

[0149] The preset time period can be considered as a time period corresponding to a certain period of time in the future from the current moment. This preset time period can be set according to historical conditions. For example, if the data change of the water meter can be clearly observed in one minute, the preset time period can be set to one minute.

[0150] Abnormal water discharge can be considered as a large difference between the water discharge volume and the usual water discharge volume. For example, the water flow rate at this moment is A, but the water discharge volume at the current moment is far more than A, indicating that there may be abnormal water discharge.

[0151] Because the present embodiment performs water meter reading through intelligent remote control, in order to determine the corresponding position of the water meter, a locator can be embedded in the water meter, and the precise positioning of the water meter can be obtained through this locator.

[0152] After obtaining this precise positioning, the corresponding outlet data can be retrieved when setting the water meter to obtain the number of outlets.

[0153] In this embodiment, each water outlet is provided with a sensor, and the sensor can sense the water flow rate of the water outlet.

[0154] When there is abnormal water discharge, the sensing result of the receiving sensor is determined to determine whether there is abnormal water discharge at the water outlet. If there is a large difference with the water discharge in the historical period, it can be considered that there is abnormal water discharge, that is, there is an abnormal water outlet. If not, it can be considered that there is a rupture inside the pipe, resulting in an increase in the total amount of water flow and abnormal water discharge.

[0155] Through the solution provided by this embodiment, by continuously taking pictures of water meter images for real-time monitoring, abnormal water discharge can be discovered in time to avoid waste and loss of water resources. If abnormal water discharge is found, the solution can obtain the precise location of the water meter, accurate to the specific water outlet, and improve the accuracy of abnormal location. By using image processing and recognition technology, abnormal water discharge can be automatically detected, avoiding the tediousness and errors of manual inspections and improving detection efficiency and accuracy.

[0156] In some embodiments, historical water flow data of a water meter is obtained; water use habits are determined based on the historical water flow data; based on the water use habits, the expected water flow at the current moment is determined; the water flow is compared with the expected water flow to determine whether there is an abnormality in the water flow; if there is an abnormality in the water flow, it is determined that there is an abnormal water outlet.

[0157] Historical water flow data can be considered as data obtained by monitoring water flow in a historical period. By analyzing this historical water flow data, the water use habits of the current location can be determined, for example, when water consumption is high and when water consumption is low.

[0158] The expected water flow can be considered as the water flow corresponding to the same time obtained by analyzing water usage habits. For example, if the water flow at 4 pm in the historical period is B, it can be predicted that if the current time is 4 pm, the water flow will also be B.

[0159] Through the solution provided by this embodiment, historical water flow data is analyzed to understand the user's water usage habits, so that the expected water flow at the current moment can be predicted more accurately. By comparing the actual water flow with the expected water flow, it is possible to more accurately determine whether the water flow is abnormal, thereby improving the accuracy of abnormality detection.

[0160] Figure 3 A schematic diagram of the structure of an AI vision-based intelligent meter reading system provided in one embodiment of the present application is shown in FIG. Figure 3 As shown, the AI ​​vision-based intelligent meter reading system 300 of this embodiment includes: a stain analysis module 301, a type determination module 302, a data distribution analysis module 303, a reading judgment module 304, and a water meter cleaning control module 305.

[0161] The stain analysis module 301 is used to obtain the water meter image taken by the camera device in real time, analyze the water meter image, and determine whether there is stain on the water meter;

[0162] A type determination module 302, configured to determine the water meter type of the water meter according to the water meter image if it is determined that there is stain;

[0163] A data distribution analysis module 303, used to determine the data distribution of the water meter according to the type of the water meter;

[0164] A reading judgment module 304 is used to determine whether the stain affects data reading according to the data distribution;

[0165] The water meter cleaning control module 305 is used to start the water meter cleaning device to clean the water meter to read the water meter value if data reading is affected.

[0166] In some embodiments, the type determination module 302 is specifically used to:

[0167] Recognize the water meter image and determine the outline of the water meter;

[0168] According to the profile, water meter data in a water meter information database is screened to determine a matching water meter;

[0169] Extracting features from the water meter image to obtain water meter features, and comparing the water meter features with features of the matching water meter;

[0170] According to the comparison result, determining whether there is a target water meter that matches all the water meter characteristics;

[0171] If it exists, the water meter information of the target water meter is retrieved to determine the water meter type of the water meter.

[0172] In some embodiments, the smart meter reading system 300 further includes a contamination type analysis 306 for:

[0173] Obtain a detail image corresponding to the stain at the data distribution position;

[0174] Binarizing the detail image to obtain a clear image, and extracting stain features of the clear image;

[0175] Clustering the stain features, analyzing the features of each cluster after clustering, and determining the characteristics of the stain;

[0176] Based on the characteristics, the type of the stain is determined.

[0177] In some embodiments, the water meter cleaning control module 305 is specifically used to:

[0178] Determine the best cleaning agent based on the type of stain in question;

[0179] Acquire the placement data of the water meter cleaning device; the placement data includes the placement quantity;

[0180] Determining whether the optimal cleaning agent exists in the water meter cleaning device according to the placement data;

[0181] If so, determining the optimal amount of cleaning agent to be used according to the size of the stain;

[0182] Comparing the used amount with the placed amount, and determining whether the stain can be completely removed according to the comparison result;

[0183] If it is determined that the stain can be completely removed, the water meter cleaning device is started to treat the stain.

[0184] In some embodiments, the smart meter reading system 300 further includes a storage quantity analysis module 307 for:

[0185] If it is determined that the stain cannot be completely removed, determining the cleaning degree of the current amount of items placed according to the cleaning data of the optimal cleaning agent;

[0186] Determining data exposure according to the cleaning degree and the data distribution;

[0187] Determining whether the stain affects meter reading according to the data revealed;

[0188] If there is no impact, start the water meter cleaning device to treat the stain.

[0189] In some embodiments, the smart meter reading system 300 further includes an abnormality analysis module 308 for:

[0190] Using the water meter images taken continuously during a preset period of time, determine whether there is currently any abnormal water discharge;

[0191] If there is water outflow abnormality at the current moment, obtain the precise location of the water meter at the current moment;

[0192] According to the precise positioning, the number of water outlets is determined, and sensors corresponding to the water outlets are activated to sense the water flow;

[0193] Determining whether there is an abnormal water outlet according to the water flow rate;

[0194] If it does not exist, it is determined that the pipeline is abnormal.

[0195] In some embodiments, the anomaly analysis module 308 is specifically used to:

[0196] Obtaining historical water flow data of the water meter;

[0197] determining water use habits based on the historical water flow data;

[0198] Determining the expected water flow at the current moment based on the water use habits;

[0199] Comparing the water flow rate with the expected water flow rate to determine whether there is an abnormality in the water flow rate;

[0200] If the water flow rate is abnormal, it is determined that there is an abnormal water outlet.

[0201] The device of this embodiment can be used to execute the method of any of the above embodiments. Its implementation principle and technical effects are similar and will not be described in detail here.

[0202] Figure 4 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application is shown in FIG. Figure 4 As shown, the electronic device 400 of this embodiment may include: a memory 401 and a processor 402 .

[0203] The memory 401 stores a computer program that can be loaded by the processor 402 and execute the method in the above embodiment.

[0204] The processor 402 and the memory 401 are connected, for example, via a bus.

[0205] Optionally, the electronic device 400 may further include a transceiver. It should be noted that in actual applications, the number of transceivers is not limited to one, and the structure of the electronic device 400 does not constitute a limitation on the embodiments of the present application.

[0206] Processor 402 may be a CPU (Central Processing Unit), a general purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 402 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0207] The bus may include a path to transmit information between the above components. The bus may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0208] The memory 401 can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited to these.

[0209] The memory 401 is used to store the application code for executing the solution of the present application, and the execution is controlled by the processor 402. The processor 402 is used to execute the application code stored in the memory 401 to implement the contents shown in the above method embodiment.

[0210] The electronic devices include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), and fixed terminals such as digital TVs, desktop computers, etc. It can also be a server, etc. Figure 4 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0211] The electronic device of this embodiment can be used to execute the method of any of the above embodiments, and its implementation principle and technical effect are similar, which will not be described in detail here.

[0212] The present application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and execute the method in the above embodiment.

[0213] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

Claims

1. An intelligent meter reading method based on AI vision, characterized in that: Applicable to an intelligent meter reading device, the intelligent meter reading device includes a water meter, a camera device, a server, and a water meter cleaning device; The method is applied to a server, and the method comprises: Acquire a water meter image captured in real time by the camera device, analyze the water meter image, and determine whether there is stain on the water meter; If it is determined that there is stain, determining the water meter type of the water meter according to the water meter image; Determining data distribution of the water meter according to the type of the water meter; Determining whether the stain affects data reading according to the data distribution; If data reading is affected, the water meter cleaning device is started to clean the water meter to read the water meter value.

2. The method according to claim 1, characterized in that The step of determining the water meter type of the water meter according to the water meter image includes: Recognize the water meter image and determine the outline of the water meter; According to the profile, water meter data in a water meter information database is screened to determine a matching water meter; Extracting features from the water meter image to obtain water meter features, and comparing the water meter features with features of the matching water meter; According to the comparison result, determining whether there is a target water meter that matches all the water meter characteristics; If it exists, the water meter information of the target water meter is retrieved to determine the water meter type of the water meter.

3. The method according to claim 1, characterized in that Before starting the water meter cleaning device to clean the water meter, the method further includes: Obtain a detail image corresponding to the stain at the data distribution position; Binarizing the detail image to obtain a clear image, and extracting stain features of the clear image; Clustering the stain features, analyzing the features of each cluster after clustering, and determining the characteristics of the stain; Based on the characteristics, the type of the stain is determined.

4. The method according to claim 3, characterized in that: The step of starting the water meter cleaning device to clean the water meter includes: Determine the best cleaning agent based on the type of stain in question; Acquire the placement data of the water meter cleaning device; the placement data includes the placement quantity; Determining whether the optimal cleaning agent exists in the water meter cleaning device according to the placement data; If so, determining the optimal amount of cleaning agent to be used according to the size of the stain; Comparing the used amount with the placed amount, and determining whether the stain can be completely removed according to the comparison result; If it is determined that the stain can be completely removed, the water meter cleaning device is started to treat the stain.

5. The method according to claim 4, characterized in that Also includes: If it is determined that the stain cannot be completely removed, determining the cleaning degree of the current amount of items placed according to the cleaning data of the optimal cleaning agent; Determining data exposure according to the cleaning degree and the data distribution; Determining whether the stain affects meter reading according to the data revealed; If there is no impact, start the water meter cleaning device to treat the stain.

6. The method according to claim 1, characterized in that Also includes: Using the water meter images taken continuously during a preset period of time, determine whether there is currently any abnormal water discharge; If there is water outflow abnormality at the current moment, obtain the precise location of the water meter at the current moment; According to the precise positioning, the number of water outlets is determined, and sensors corresponding to the water outlets are activated to sense the water flow; Determining whether there is an abnormal water outlet according to the water flow rate; If it does not exist, it is determined that the pipeline is abnormal.

7. The method according to claim 6, characterized in that The determining whether there is an abnormal water outlet according to the water flow rate includes: Obtaining historical water flow data of the water meter; determining water use habits based on the historical water flow data; Determining the expected water flow at the current moment based on the water use habits; Comparing the water flow rate with the expected water flow rate to determine whether there is an abnormality in the water flow rate; If the water flow rate is abnormal, it is determined that there is an abnormal water outlet.

8. An AI vision-based intelligent meter reading system, characterized in that: include: A stain analysis module, used to obtain the water meter image taken by the camera device in real time, analyze the water meter image, and determine whether there is stain on the water meter; a type determination module, configured to determine the water meter type of the water meter according to the water meter image if it is determined that there is stain; A data distribution analysis module, used to determine the data distribution of the water meter according to the type of the water meter; A reading judgment module, used to determine whether the stain affects data reading according to the data distribution; The water meter cleaning control module is used to start the water meter cleaning device to clean the water meter to read the water meter value if the data reading is affected.

9. An electronic device, characterized in that: include: Memory and processor; The memory is used to store program instructions; The processor is used to call and execute the program instructions in the memory to perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images