A smart meter automatic inspection system and method
Through the smart meter automatic patrol system, combined with image recognition and automated testing equipment, the problem of inefficiency of existing detection methods is solved, and efficient and accurate detection of smart meters is achieved to meet the needs of large-scale deployment.
Patent Information
- Application Number
- CN202411178944.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-27
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-08-27
AI Technical Summary
The existing smart meter detection methods are inefficient and are susceptible to human factors, making it difficult to achieve batch efficient inspection, and cannot meet the needs of large-scale deployment and rapid inspection of smart meters.
The intelligent meter automatic inspection system is adopted, combined with image recognition technology and automated testing equipment, and through power-on testing, water vapor pressure detection and image analysis, the efficient and accurate detection of smart meter is achieved.
It improves the reliability and efficiency of the detection results, can detect multiple smart meters at one time, reduces manual operations, ensures the accuracy of the acquisition points and the accuracy of the detection results, and meets the needs of large-scale deployment.
Smart Images

Figure CN118707435B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of smart meter detection, and in particular to a smart meter automatic inspection system and method. Background Art
[0002] With the development of smart grid technology, smart meters have become an important part of modern power systems. Smart meters can not only accurately record electricity consumption, but also have functions such as remote communication and real-time monitoring, which greatly improves the efficiency and management level of the power grid. However, the quality and performance of smart meters directly affect the operating stability and user experience of the power system. Therefore, it is particularly important to conduct strict testing of smart meters.
[0003] In the prior art, there are some problems with the detection methods of smart meters. The traditional methods mainly rely on manual operation and visual inspection, which is not only inefficient but also easily affected by human factors, resulting in inaccurate detection results. Especially when facing a large number of meters, the detection personnel are prone to fatigue, which further reduces the detection accuracy and reliability. In addition, when facing a large number of smart meters, the traditional testing methods are difficult to achieve efficient batch detection and cannot meet the needs of large-scale deployment and rapid detection of smart meters. Currently, a more advanced method for detecting the air tightness of smart meters is to use an air tightness detection instrument for detection, which mainly places the smart meter in a sealable detection device and introduces gas into the device. The air tightness of the smart meter is detected according to the introduced gas and pressure changes. This detection method generally places only one smart meter in the device at a time, and the detection efficiency is relatively low.
[0004] This scheme proposes an automatic inspection method for smart meters, which can achieve efficient and accurate detection of smart meters by combining image recognition technology and automated testing equipment. Summary of the invention
[0005] The present invention provides a smart meter automatic inspection system and method, which facilitates solving the problems mentioned in the above background technology.
[0006] The present invention provides the following technical solution: a method for automatic inspection of a smart meter, comprising:
[0007] Move the smart meter to the test location and conduct a power-on test on the smart meter;
[0008] During the power-on test, the image of the smart meter display screen is collected and recorded as the initial image;
[0009] According to the text distribution in the initial image, obtain and mark the acquisition points;
[0010] Acquire a display screen image of a smart meter at a collection point, and record the obtained image as a first analysis image;
[0011] After obtaining the collection points, the smart meter rules are placed in the test container;
[0012] Seal the test container and introduce water vapor into the test container;
[0013] Monitor the internal pressure of the test container. When the pressure is equal to the preset pressure, stop introducing water vapor and obtain the internal temperature of the test container at this time.
[0014] Maintaining the pressure inside the test container and heating the test container to keep the temperature inside the test container constant;
[0015] When the time for maintaining the internal pressure and temperature of the test container reaches a set threshold, the heating of the test container is stopped;
[0016] Remove the smart meter from the test container;
[0017] Cool down the smart meter at the collection point;
[0018] After the cooling operation is completed, remove the condensed water on the outer surface of the smart meter;
[0019] Acquire the smart meter display screen image at the collection point again, and record the obtained image as the second analysis image;
[0020] The first analysis image is compared with the second analysis image to determine whether to mark the smart meter.
[0021] Optionally, obtaining the acquisition points according to the text distribution in the initial image specifically includes:
[0022] Get the geometric center point of each character in the initial image and record it as the character center point;
[0023] Calculate the distance between the center point of each character and the center points of other characters;
[0024] Set the distance threshold A;
[0025] For each text center point, count the number of other text center points whose distance to the text center point is within A, and record it as the number of dense points;
[0026] Get the number of dense points at the center of each character and compare them;
[0027] Select the center point of the text with the largest number of dense points and record it as the collection point to be determined;
[0028] If there is only one collection point to be confirmed, then the collection point to be confirmed is used as the collection point;
[0029] If there are multiple collection points to be confirmed, the distance between every two collection points to be confirmed is obtained, and the average distance is calculated;
[0030] Draw a circle with each collection point to be confirmed as the center and the average distance as the radius;
[0031] Obtain the number of unconfirmed collection points in each circle respectively, and record it as the dense number;
[0032] Get the circle with the largest number of dense points and mark the collection points to be confirmed in the circle;
[0033] Get the center point between the marked collection points to be confirmed, and use the center point as the collection point.
[0034] Optionally, the step of acquiring the image of the smart meter display screen at the collection point and recording the obtained image as the first analysis image is as follows:
[0035] Draw a vertical line through the collection point perpendicular to the display screen of the smart meter, which is recorded as the reference vertical line;
[0036] Taking the collection point as the starting point, on the plane where the smart meter display screen is located, 8 rays are drawn at intervals of 45 degrees, which are recorded as ray 1, ray 2, ray 3, ray 4, ray 5, ray 6, ray 7 and ray 8 in sequence;
[0037] Obtain a straight line that equally divides the angle between the first ray and the reference vertical line, and record it as the first angle line;
[0038] Obtain a straight line that equally divides the angle between the second ray and the reference vertical line, and record it as the second angle line;
[0039] Obtain a straight line that equally divides the angle between the third ray and the reference vertical line, and record it as the third angle line;
[0040] Obtain a straight line that equally divides the angle between the fourth ray and the reference vertical line, and record it as the fourth angle line;
[0041] Obtain a straight line that equally divides the angle between the fifth ray and the reference vertical line, and record it as the fifth angle line;
[0042] Obtain a straight line that equally divides the angle between the sixth ray and the reference vertical line, and record it as the sixth angle line;
[0043] Obtain a straight line that equally divides the angle between the seventh ray and the reference vertical line, and record it as the seventh angle line;
[0044] Obtain a straight line that equally divides the angle between the eighth ray and the reference vertical line, and record it as the eighth angle line;
[0045] Nine directions pointing to the collection point along angle line No. 1, angle line No. 2, angle line No. 3, angle line No. 4, angle line No. 5, angle line No. 6, angle line No. 7, angle line No. 8 and the reference vertical line are used as nine image collection directions;
[0046] The smart meter display screen images are captured in nine image capture directions respectively, and the obtained images are recorded as the first analysis images.
[0047] Optionally, after obtaining the collection point, placing the smart meter rule in a test container includes:
[0048] Get the bottom circle of the test container, draw a straight line tangent to the bottom circle, and record it as reference straight line No. 1;
[0049] Draw a straight line perpendicular to reference line No. 1 and tangent to the bottom circle, record it as reference line No. 2;
[0050] Get the length and width of the bottom of the smart meter, denoted as B and C respectively;
[0051] Set the interval distance D;
[0052] On the bottom circle, starting from reference line 1, draw a straight line perpendicular to reference line 2 and parallel to reference line 1 every B+D.
[0053] On the bottom circle, starting from reference line No. 2, draw a straight line perpendicular to reference line No. 1 and parallel to reference line No. 2 every C+D;
[0054] The straight lines divide the bottom circle into several rectangular areas;
[0055] Get all the rectangular areas that are completely on the bottom circle and record them as selected areas;
[0056] Get the center point of the selected area and record it as the placement point.
[0057] Optionally, after obtaining the collection point, placing the smart meter rule in a test container further includes:
[0058] Obtain the center point of the bottom surface of the smart meter and record it as the meter center;
[0059] Get the number of placement points in the test container, recorded as E;
[0060] Select a number E of smart meters and place them in the test container in sequence;
[0061] Each time a placement is performed, a placement point is selected and the meter center of the placed smart meter is made to coincide with the placement point.
[0062] Optionally, the pressure inside the test container is monitored, and when the pressure is equal to a preset pressure, the introduction of water vapor is stopped, and the temperature inside the test container at this time is obtained, specifically:
[0063] Set the preset pressure, denoted as F;
[0064] During the process of introducing water vapor into the test container, the internal pressure value of the test container is monitored;
[0065] When the internal pressure value of the test container is monitored to be equal to F, stop introducing water vapor and keep the test container in a sealed state;
[0066] Measure the current temperature inside the sealed container, recorded as G;
[0067] The method of maintaining the pressure inside the test container and heating the test container to make the temperature inside the test container constant is specifically as follows:
[0068] heating the test container and monitoring the temperature inside the test container;
[0069] Maintain the internal temperature of the test container at G;
[0070] When the time for maintaining the internal pressure and temperature of the test container reaches a set threshold, the heating of the test container is stopped, specifically:
[0071] Set the insulation time H;
[0072] Start timing by heating the test container;
[0073] When the time duration is equal to H, stop heating;
[0074] The smart meter in the test container is cooled naturally at room temperature.
[0075] Optionally, acquiring the smart meter display screen image at the collection point again and recording the obtained image as a second analysis image includes:
[0076] For the smart meter with condensed water on the outer surface removed, images of the smart meter display screen are captured again in nine image capture directions, and the obtained images are recorded as second analysis images.
[0077] Optionally, comparing the first analysis image with the second analysis image to determine whether to mark the smart meter includes:
[0078] dividing the first analysis image and the second analysis image acquired in the same image acquisition direction into a group of analysis images;
[0079] comparing a first analysis image and a second analysis image in each group of analysis images by image recognition;
[0080] Set the similarity standard K;
[0081] If the similarity between the first analysis image and the second analysis image in the same group of analysis images is greater than or equal to K, then the group of analysis images is recorded as a similar group;
[0082] If the similarity between the first analysis image and the second analysis image in the same group of analysis images is less than K, the group of analysis images is recorded as a deviation group;
[0083] If all nine groups of analyzed images are similar groups, the smart meter will not be marked;
[0084] If there is a group of analysis images that is a deviation group, the smart meters corresponding to the nine groups of analysis images are marked.
[0085] A system for realizing automatic inspection of the smart meter, comprising:
[0086] Transport module: used to transport smart meters and transfer them to different locations;
[0087] Image acquisition module: used to acquire the image of the smart meter display screen before and after the test;
[0088] Image recognition module: analyzes the acquired images and compares the similarity between two analyzed images;
[0089] Distance measurement module: obtain the distance between the center points of the text;
[0090] Test container: The bottom surface is round and sealed when closed, used to hold smart meters;
[0091] Pressure acquisition module: used to monitor the internal pressure of the test container;
[0092] Temperature acquisition module: used to monitor the internal temperature of the test container;
[0093] Heating module: used to heat the test container and keep the internal temperature of the test container constant;
[0094] Cooling module: used to cool the display screen;
[0095] Data analysis module: compares similarity information and determines whether to mark the smart meter.
[0096] The present invention has the following beneficial effects:
[0097] 1. Automatically obtain and mark the collection points according to the text distribution in the initial image to ensure the accuracy of the collection point selection. This method can avoid inconsistent collection point selection due to human judgment and improve the reliability of the detection results. By automatically analyzing the text distribution, the system can quickly and accurately mark the collection points, reducing the time and cost of manual operation.
[0098] 2. Through regular placement, the space in the test container is fully utilized, so that more smart meters can be placed at one time, increasing the number of smart meters that can be tested at a single time, thereby improving the detection efficiency.
[0099] 3. By removing the condensed water on the outer surface, the interference of the condensed water on the outer surface on the test results can be effectively avoided, the accuracy of image acquisition can be improved, the clarity and accuracy of subsequent image acquisition can be ensured, and a reliable data basis can be provided for comparative analysis.
[0100] 4. By introducing water vapor into a closed container and making the container have a high pressure, when the smart meter in the container has poor airtightness, water vapor enters the smart meter due to pressure, and then the smart meter is cooled to condense the water vapor entering the smart meter into water. The airtightness of the smart meter is judged by detecting whether condensed water is detected in the smart meter. This method can detect multiple smart meters at one time, and then detect the airtightness of each smart meter by performing image recognition on each smart meter, and can separate smart meters with poor airtightness. This detection method can improve the efficiency of airtightness detection of smart meters.
[0101] 5. By calculating the number of dense points at the center of each character and selecting the center of the character with the largest number of dense points as the collection point, the rationality and accuracy of the collection point are ensured, thereby further improving the comprehensiveness and reliability of the detection.
[0102] 6. By setting nine image acquisition directions, the image acquisition of the smart meter is completed, ensuring the all-round detection of the smart meter display screen. This all-round, multi-angle detection method can effectively capture all information on the smart meter display screen and avoid omissions or errors that may be caused by a single angle. Through multi-angle analysis, the accuracy and comprehensiveness of the detection results can be improved, ensuring that any subtle changes can be captured, which can improve the reliability of detection and provide multi-dimensional data support for product quality control.
[0103] 7. By collecting images of smart meters at collection points, which are the locations where text is most concentrated on the display screen of smart meters, collecting images at these locations can make the collected images more representative and comprehensive, thus ensuring the accuracy of the test results.
[0104] 8. Acquiring the second analysis image at the same position and direction can reduce the error caused by position and angle changes during the detection process, ensure the uniformity and consistency of the comparison images, and enhance the credibility of the detection results.
[0105] 9. By comparing the image collected with the area with the densest text, it is possible to detect whether there is condensation inside the smart meter. If there is condensation inside the smart meter, the contrast will be more obvious when observed at the location with the densest text. The dense text area provides a background with high contrast. When there is condensation at this location, the image change is more obvious and can be more clearly reflected in the image. This method can reduce detection errors and improve the accuracy of detection results. BRIEF DESCRIPTION OF THE DRAWINGS
[0106] Figure 1 A schematic diagram of the rays on the display screen of the smart electric meter of the present invention.
[0107] Figure 2 This is a schematic diagram of angle line No. 1 of the present invention.
[0108] Figure 3 Schematic diagram of the selected area for the present invention. DETAILED DESCRIPTION
[0109] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0110] Embodiment 1, a method for automatic inspection of a smart meter, comprising:
[0111] Move the smart meter to the test location and conduct a power-on test on the smart meter;
[0112] During the power-on test, the image of the smart meter display screen is collected and recorded as the initial image;
[0113] According to the distribution of text in the initial image, the collection points are obtained and marked; according to the distribution of text in the initial image, the collection points are automatically obtained and marked to ensure that the selection of collection points is objective and fair. This method can avoid inconsistent collection point selection due to human judgment and improve the reliability of the detection results. By automatically analyzing the text distribution, the system can quickly and accurately mark the collection points, reducing the time and cost of manual operation.
[0114] Acquire a display screen image of a smart meter at a collection point, and record the obtained image as a first analysis image;
[0115] After obtaining the collection points, the smart meters are placed in the test container in a regular manner. By placing the smart meters in a regular manner, the space in the test container is fully utilized, so that more smart meters can be placed at one time, the number of smart meters that can be detected at one time is increased, and the detection efficiency is thereby improved.
[0116] Seal the test container and introduce water vapor into the test container;
[0117] Monitor the internal pressure of the test container. When the pressure is equal to the preset pressure, stop introducing water vapor and obtain the internal temperature of the test container at this time.
[0118] Maintaining the pressure inside the test container and heating the test container to keep the temperature inside the test container constant;
[0119] When the time for maintaining the internal pressure and temperature of the test container reaches a set threshold, the heating of the test container is stopped;
[0120] Remove the smart meter from the test container;
[0121] Cool down the smart meter at the collection point;
[0122] After completing the cooling operation, remove the condensed water on the outer surface of the smart meter. By removing the condensed water on the outer surface, the interference of the condensed water on the outer surface on the detection results can be effectively avoided, the accuracy of image acquisition can be improved, the clarity and accuracy of subsequent image acquisition can be ensured, and a reliable data basis can be provided for comparative analysis.
[0123] Acquire the smart meter display screen image at the collection point again, and record the obtained image as the second analysis image;
[0124] Compare the first analysis image with the second analysis image to determine whether to mark the smart meter. By introducing water vapor into a sealed container and making the container have a large pressure, when the airtightness of the smart meter in the container is poor, the water vapor enters the smart meter due to the pressure, and then the smart meter is cooled to condense the water vapor entering the smart meter into water. The airtightness of the smart meter is determined by detecting whether condensed water is detected in the smart meter. This method can detect multiple smart meters at one time, and then detect the airtightness of each smart meter by performing image recognition on each smart meter, and can separate smart meters with poor airtightness. This detection method can improve the efficiency of airtightness detection of smart meters.
[0125] The step of obtaining the acquisition points according to the text distribution in the initial image specifically includes:
[0126] Get the geometric center point of each character in the initial image and record it as the character center point;
[0127] Calculate the distance between the center point of each character and the center points of other characters;
[0128] Set the distance threshold A; the distance threshold A can be set according to the distance between each character center point and other character center points, and the average value of all distances can be selected as the distance threshold A.
[0129] For each text center point, count the number of other text center points whose distance to the text center point is within A, and record it as the number of dense points;
[0130] Get the number of dense points at the center of each character and compare them;
[0131] Select the center point of the text with the largest number of dense points and record it as the collection point to be determined;
[0132] If there is only one collection point to be confirmed, then the collection point to be confirmed is used as the collection point;
[0133] If there are multiple collection points to be confirmed, the distance between every two collection points to be confirmed is obtained, and the average distance is calculated;
[0134] Draw a circle with each collection point to be confirmed as the center and the average distance as the radius;
[0135] Obtain the number of unconfirmed collection points in each circle respectively, and record it as the dense number;
[0136] Get the circle with the largest number of dense points and mark the collection points to be confirmed in the circle;
[0137] Obtain the center point between the marked collection points to be confirmed and use it as the collection point. By calculating the number of dense points at each text center point and selecting the text center point with the largest number of dense points as the collection point, the rationality and accuracy of the collection point are ensured, thereby further improving the comprehensiveness and reliability of the detection.
[0138] The smart meter display screen image is obtained at the collection point, and the obtained image is recorded as the first analysis image, specifically:
[0139] Draw a vertical line through the collection point perpendicular to the display screen of the smart meter, which is recorded as the reference vertical line;
[0140] Reference Figure 1, taking the collection point as the starting point, on the plane where the smart meter display screen is located, 8 rays are made at an interval of 45 degrees, which are successively recorded as ray No. 1, ray No. 2, ray No. 3, ray No. 4, ray No. 5, ray No. 6, ray No. 7 and ray No. 8; the 45-degree interval can be changed accordingly according to actual production needs. The smaller the set angle, the more images are collected, the more refined the state of the smart meter can be detected, and the accuracy of detection is improved, but the time spent on collecting images will be increased accordingly.
[0141] Reference Figure 2 , obtain the straight line that equally divides the angle between the No. 1 ray and the reference vertical line, recorded as the No. 1 angle line;
[0142] Obtain a straight line that equally divides the angle between the second ray and the reference vertical line, and record it as the second angle line;
[0143] Obtain a straight line that equally divides the angle between the third ray and the reference vertical line, and record it as the third angle line;
[0144] Obtain a straight line that equally divides the angle between the fourth ray and the reference vertical line, and record it as the fourth angle line;
[0145] Obtain a straight line that equally divides the angle between the fifth ray and the reference vertical line, and record it as the fifth angle line;
[0146] Obtain a straight line that equally divides the angle between the sixth ray and the reference vertical line, and record it as the sixth angle line;
[0147] Obtain a straight line that equally divides the angle between the seventh ray and the reference vertical line, and record it as the seventh angle line;
[0148] Obtain a straight line that equally divides the angle between the eighth ray and the reference vertical line, and record it as the eighth angle line;
[0149] Nine directions pointing to the collection point along angle line No. 1, angle line No. 2, angle line No. 3, angle line No. 4, angle line No. 5, angle line No. 6, angle line No. 7, angle line No. 8 and the reference vertical line are used as nine image collection directions; by setting nine image collection directions, the collection of smart meter images is completed, ensuring all-round detection of the smart meter display screen. This all-round, multi-angle detection method can effectively capture all information on the smart meter display screen and avoid omissions or errors that may be caused by a single angle. Through multi-angle analysis, the accuracy and comprehensiveness of the detection results can be improved to ensure that any subtle changes can be captured, which can improve the reliability of detection and provide multi-dimensional data support for product quality control.
[0150] The images of the smart meter display screen are collected in nine image collection directions respectively, and the obtained images are recorded as the first analysis images. By collecting the smart meter images at the collection points, which are the locations where the text displayed on the text smart meter display screen is most concentrated, collecting images at this location can make the collected images more representative and comprehensive, thus ensuring the accuracy of the detection results.
[0151] After obtaining the collection point, the smart meter rules are placed in the test container, including:
[0152] Reference Figure 3 , obtain the bottom circle of the test container, and draw a straight line tangent to the bottom circle, which is recorded as reference straight line No. 1;
[0153] Draw a straight line perpendicular to reference line No. 1 and tangent to the bottom circle, record it as reference line No. 2;
[0154] Get the length and width of the bottom of the smart meter, denoted as B and C respectively;
[0155] Set the interval distance D; the interval distance D is the interval between two smart meters, and the value of D can be set arbitrarily according to detection needs.
[0156] On the bottom circle, starting from reference line 1, draw a straight line perpendicular to reference line 2 and parallel to reference line 1 every B+D.
[0157] On the bottom circle, starting from reference line No. 2, draw a straight line perpendicular to reference line No. 1 and parallel to reference line No. 2 every C+D;
[0158] The straight lines divide the bottom circle into several rectangular areas;
[0159] Get all the rectangular areas that are completely on the bottom circle and record them as the selected areas, such as Figure 3 A rectangular area marked in ;
[0160] Get the center point of the selected area and record it as the placement point.
[0161] After obtaining the collection point, placing the smart meter rule in the test container also includes:
[0162] Obtain the center point of the bottom surface of the smart meter and record it as the meter center;
[0163] Get the number of placement points in the test container, recorded as E;
[0164] Select a number E of smart meters and place them in the test container in sequence;
[0165] Each time a placement is performed, a placement point is selected and the meter center of the placed smart meter is made to coincide with the placement point.
[0166] The internal pressure of the test container is monitored, and when the pressure is equal to the preset pressure, the introduction of water vapor is stopped, and the internal temperature of the test container at this time is obtained, specifically:
[0167] A preset pressure is set, denoted as F. The setting of F needs to take into account the pressure bearing capacity of the smart meter. Under the pressure bearing capacity of the smart meter, the larger the F setting is, the faster water vapor can enter the smart meter with poor airtightness, thereby improving the detection efficiency.
[0168] During the process of introducing water vapor into the test container, the internal pressure value of the test container is monitored;
[0169] When the internal pressure value of the test container is monitored to be equal to F, stop introducing water vapor and keep the test container in a sealed state;
[0170] Measure the current temperature inside the sealed container, recorded as G;
[0171] The method of maintaining the pressure inside the test container and heating the test container to make the temperature inside the test container constant is specifically as follows:
[0172] heating the test container and monitoring the temperature inside the test container;
[0173] Maintain the internal temperature of the test container at G;
[0174] When the time for maintaining the internal pressure and temperature of the test container reaches a set threshold, the heating of the test container is stopped, specifically:
[0175] Set the insulation time H; the value of H is set according to the actual detection needs. The smaller H is set, the less time it takes to detect. However, it may cause water vapor to not have enough time to enter the smart meter with poor airtightness, thereby affecting the accuracy of the detection.
[0176] Start timing by heating the test container;
[0177] When the time duration is equal to H, stop heating;
[0178] The smart meter in the test container is cooled naturally at room temperature.
[0179] The step of acquiring the smart meter display screen image at the collection point again and recording the obtained image as a second analysis image includes:
[0180] For the smart meter with condensed water removed from the outer surface, the display screen image of the smart meter is collected again in nine image collection directions, and the obtained image is recorded as the second analysis image. By obtaining the second analysis image in the same position and direction, the error caused by position and angle changes during the detection process can be reduced, the uniformity and consistency of the comparison image can be ensured, and the credibility of the detection result can be enhanced.
[0181] The comparing the first analysis image with the second analysis image to determine whether to mark the smart meter includes:
[0182] dividing the first analysis image and the second analysis image acquired in the same image acquisition direction into a group of analysis images;
[0183] The first analysis image and the second analysis image in each group of analysis images are compared through image recognition; by comparing with the image collected in the area with the densest text, it is detected whether there is condensation inside the smart meter. If there is condensation inside the smart meter, the contrast is more obvious when observed at the position with the densest text. The dense text area provides a background with high contrast. When there is condensation at this position, the image change is more obvious and can be more clearly reflected in the image. This method can reduce detection errors and improve the accuracy of detection results.
[0184] A similarity standard K is set; the K value reflects the similarity between the first analysis image and the second analysis image. The larger the K value is, the more similar the first analysis image and the second analysis image are required to be. If the K value is set to 1, the first analysis image and the second analysis image are required to be completely similar and the two images are completely consistent.
[0185] If the similarity between the first analysis image and the second analysis image in the same group of analysis images is greater than or equal to K, then the group of analysis images is recorded as a similar group;
[0186] If the similarity between the first analysis image and the second analysis image in the same group of analysis images is less than K, the group of analysis images is recorded as a deviation group;
[0187] If all nine groups of analyzed images are similar groups, the smart meter will not be marked; this indicates that the smart meter has good air tightness and no condensation water is detected inside.
[0188] If there is a group of analysis images that is a deviation group, the smart meters corresponding to the nine groups of analysis images are marked. The marked smart meters are products with poor airtightness and need to be eliminated.
[0189] Embodiment 2, a system for implementing the automatic inspection method of a smart meter, comprising:
[0190] Transport module: used to transport smart meters and transfer them to different locations;
[0191] Image acquisition module: used to acquire the image of the smart meter display screen before and after the test;
[0192] Image recognition module: analyzes the acquired images and compares the similarity between two analyzed images;
[0193] Distance measurement module: obtain the distance between the center points of the text;
[0194] Test container: The bottom surface is round and sealed when closed, used to hold smart meters;
[0195] Pressure acquisition module: used to monitor the internal pressure of the test container;
[0196] Temperature acquisition module: used to monitor the internal temperature of the test container;
[0197] Heating module: used to heat the test container and keep the internal temperature of the test container constant;
[0198] Cooling module: used to cool the display screen;
[0199] Data analysis module: compares similarity information and determines whether to mark the smart meter.
[0200] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0201] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
Claims
1. A method for automatic inspection of smart meters. It is characterized in that include: Move the smart meter to the test location and conduct a power-on test on the smart meter; During the power-on test, the image of the smart meter display screen is collected and recorded as the initial image; According to the text distribution in the initial image, obtain and mark the acquisition points; Acquire a display screen image of a smart meter at a collection point, and record the obtained image as a first analysis image; After obtaining the collection points, the smart meter rules are placed in the test container; Seal the test container and introduce water vapor into the test container; Monitor the internal pressure of the test container. When the pressure is equal to the preset pressure, stop introducing water vapor and obtain the internal temperature of the test container at this time. Maintaining the pressure inside the test container and heating the test container to keep the temperature inside the test container constant; When the time for maintaining the internal pressure and temperature of the test container reaches a set threshold, the heating of the test container is stopped; Remove the smart meter from the test container; Cool down the smart meter at the collection point; After the cooling operation is completed, remove the condensed water on the outer surface of the smart meter; Acquire the smart meter display screen image at the collection point again, and record the obtained image as the second analysis image; Comparing the first analysis image with the second analysis image, and determining whether to mark the smart meter; The step of obtaining the acquisition points according to the text distribution in the initial image specifically includes: Get the geometric center point of each character in the initial image and record it as the character center point; Calculate the distance between the center point of each character and the center points of other characters; Set the distance threshold A; For each text center point, count the number of other text center points whose distance to the text center point is within A, and record it as the number of dense points; Get the number of dense points at the center of each character and compare them; Select the center point of the text with the largest number of dense points and record it as the collection point to be determined; If there is only one collection point to be confirmed, then the collection point to be confirmed is used as the collection point; If there are multiple collection points to be confirmed, the distance between every two collection points to be confirmed is obtained, and the average distance is calculated; Draw a circle with each collection point to be confirmed as the center and the average distance as the radius; Obtain the number of unconfirmed collection points in each circle respectively, and record it as the dense number; Get the circle with the largest number of dense points and mark the collection points to be confirmed in the circle; Obtain the center point between the marked collection points to be confirmed, and use the center point as the collection point; The smart meter display screen image is obtained at the collection point, and the obtained image is recorded as the first analysis image, specifically: Draw a vertical line through the collection point perpendicular to the display screen of the smart meter, which is recorded as the reference vertical line; Taking the collection point as the starting point, on the plane where the smart meter display screen is located, 8 rays are drawn at intervals of 45 degrees, which are recorded as ray 1, ray 2, ray 3, ray 4, ray 5, ray 6, ray 7 and ray 8 in sequence; Obtain a straight line that equally divides the angle between the first ray and the reference vertical line, and record it as the first angle line; Obtain a straight line that equally divides the angle between the second ray and the reference vertical line, and record it as the second angle line; Obtain a straight line that equally divides the angle between the third ray and the reference vertical line, and record it as the third angle line; Obtain a straight line that equally divides the angle between the fourth ray and the reference vertical line, and record it as the fourth angle line; Obtain a straight line that equally divides the angle between the fifth ray and the reference vertical line, and record it as the fifth angle line; Obtain a straight line that equally divides the angle between the sixth ray and the reference vertical line, and record it as the sixth angle line; Obtain a straight line that equally divides the angle between the seventh ray and the reference vertical line, and record it as the seventh angle line; Obtain a straight line that equally divides the angle between the eighth ray and the reference vertical line, and record it as the eighth angle line; Nine directions pointing to the collection point along angle line No. 1, angle line No. 2, angle line No. 3, angle line No. 4, angle line No. 5, angle line No. 6, angle line No. 7, angle line No. 8 and the reference vertical line are used as nine image collection directions; The smart meter display screen images are captured in nine image capture directions respectively, and the obtained images are recorded as the first analysis images.
2. According to claim 1, a smart meter automatic inspection method, It is characterized in that After obtaining the collection point, the smart meter rules are placed in the test container, including: Get the bottom circle of the test container, draw a straight line tangent to the bottom circle, and record it as reference straight line No. 1; Draw a straight line perpendicular to reference line No. 1 and tangent to the bottom circle, record it as reference line No. 2; Get the length and width of the bottom of the smart meter, denoted as B and C respectively; Set the interval distance D; On the bottom circle, starting from reference line 1, draw a straight line perpendicular to reference line 2 and parallel to reference line 1 every B+D. On the bottom circle, starting from reference line No. 2, draw a straight line perpendicular to reference line No. 1 and parallel to reference line No. 2 every C+D; The straight lines divide the bottom circle into several rectangular areas; Get all the rectangular areas that are completely on the bottom circle and record them as selected areas; Get the center point of the selected area and record it as the placement point.
3. According to claim 1, a smart meter automatic inspection method, It is characterized in that After obtaining the collection point, placing the smart meter rule in the test container also includes: Obtain the center point of the bottom surface of the smart meter and record it as the meter center; Get the number of placement points in the test container, recorded as E; Select a number E of smart meters and place them in the test container in sequence; Each time a placement is performed, a placement point is selected and the meter center of the placed smart meter is made to coincide with the placement point.
4. The automatic inspection method for smart electric meters according to claim 1, It is characterized in that The internal pressure of the test container is monitored, and when the pressure is equal to the preset pressure, the introduction of water vapor is stopped, and the internal temperature of the test container at this time is obtained, specifically: Set the preset pressure, denoted as F; During the process of introducing water vapor into the test container, the internal pressure value of the test container is monitored; When the internal pressure value of the test container is monitored to be equal to F, stop introducing water vapor and keep the test container in a sealed state; Measure the current temperature inside the sealed container, recorded as G; The method of maintaining the pressure inside the test container and heating the test container to make the temperature inside the test container constant is specifically as follows: heating the test container and monitoring the temperature inside the test container; Maintain the internal temperature of the test container at G; When the time for maintaining the internal pressure and temperature of the test container reaches a set threshold, the heating of the test container is stopped, specifically: Set the insulation time H; Start timing by heating the test container; When the time duration is equal to H, stop heating; The smart meter in the test container is cooled naturally at room temperature.
5. The automatic inspection method for smart meters according to claim 1, It is characterized in that The step of acquiring the smart meter display screen image at the collection point again and recording the obtained image as a second analysis image includes: For the smart meter with condensed water on the outer surface removed, images of the smart meter display screen are captured again in nine image capture directions, and the obtained images are recorded as second analysis images.
6. The automatic inspection method for smart electric meters according to claim 1, It is characterized in that The comparing the first analysis image with the second analysis image to determine whether to mark the smart meter includes: dividing the first analysis image and the second analysis image acquired in the same image acquisition direction into a group of analysis images; comparing a first analysis image and a second analysis image in each group of analysis images by image recognition; Set the similarity standard K; If the similarity between the first analysis image and the second analysis image in the same group of analysis images is greater than or equal to K, then the group of analysis images is recorded as a similar group; If the similarity between the first analysis image and the second analysis image in the same group of analysis images is less than K, the group of analysis images is recorded as a deviation group; If all nine groups of analyzed images are similar groups, the smart meter will not be marked; If there is a group of analysis images that is a deviation group, the smart meters corresponding to the nine groups of analysis images are marked.
7. A system using the smart meter automatic inspection method according to claim 1, It is characterized in that include: Transport module: used to transport smart meters and transfer them to different locations; Image acquisition module: used to acquire the image of the smart meter display screen before and after the test; Image recognition module: analyzes the acquired images and compares the similarity between two analyzed images; Distance measurement module: obtain the distance between the center points of the text; Test container: The bottom surface is round and sealed when closed, used to hold smart meters; Pressure acquisition module: used to monitor the internal pressure of the test container; Temperature acquisition module: used to monitor the internal temperature of the test container; Heating module: used to heat the test container and keep the internal temperature of the test container constant; Cooling module: used to cool the display screen; Data analysis module: compares similarity information and determines whether to mark the smart meter.
Citation Information
Patent Citations
Text line extraction method and device
CN109657629A
Pressure bearing detection method, device, control cabinet and system of base meter
CN115855211A
Electric energy meter assembling process detection device and method based on deep learning image recognition
CN116893386A
Method, device and equipment for detecting display defects of broken code screen electric meter and medium
CN117169806A
Measurement of the leak-tightness of a sealed container, e.g. a water- proof watch, by application of an overpressure to the container within a pressurizing chamber and then measurement of its resultant deformation
DE102004012534A1
