CN114926841B - Electronic water meter reading image identification method and device based on improved threading method - Google Patents
Electronic water meter reading image identification method and device based on improved threading method Download PDFInfo
- Publication number
- CN114926841B CN114926841B CN202210861687.2A CN202210861687A CN114926841B CN 114926841 B CN114926841 B CN 114926841B CN 202210861687 A CN202210861687 A CN 202210861687A CN 114926841 B CN114926841 B CN 114926841B
- Authority
- CN
- China
- Prior art keywords
- water meter
- image
- scanning
- pixel value
- threading method
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/146—Aligning or centring of the image pick-up or image-field
- G06V30/147—Determination of region of interest
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01F—MEASURING VOLUME, VOLUME FLOW, MASS FLOW OR LIQUID LEVEL; METERING BY VOLUME
- G01F15/00—Details of, or accessories for, apparatus of groups G01F1/00 - G01F13/00 insofar as such details or appliances are not adapted to particular types of such apparatus
- G01F15/06—Indicating or recording devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/146—Aligning or centring of the image pick-up or image-field
- G06V30/1475—Inclination or skew detection or correction of characters or of image to be recognised
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q9/00—Arrangements in telecontrol or telemetry systems for selectively calling a substation from a main station, in which substation desired apparatus is selected for applying a control signal thereto or for obtaining measured values therefrom
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q2209/00—Arrangements in telecontrol or telemetry systems
- H04Q2209/60—Arrangements in telecontrol or telemetry systems for transmitting utility meters data, i.e. transmission of data from the reader of the utility meter
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Fluid Mechanics (AREA)
- Arrangements For Transmission Of Measured Signals (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an electronic water meter reading image recognition method and device based on an improved threading method, wherein the method comprises the steps of inputting a reading image of an electronic water meter, and recognizing the reading number of the water meter in the image by adopting the improved threading method, wherein the improved threading method comprises the following steps: scanning from top to bottom from the midpoint of the upper boundary of the number, scanning from left to right from the 3/4 midpoint of the left boundary of the number and the 1/4 midpoint of the left boundary of the number respectively, and determining the number type according to the results of three times of scanning to obtain the accumulated flow indicating value of the detected water meter. In the traditional threading method identification mode, 7 threads need to be scanned, so that the identification efficiency is relatively low; the improved threading method only needs to scan 3 lines, so that the image recognition algorithm is greatly simplified, the running time of the image recognition algorithm is shortened, the automatic reading efficiency of the water meter is improved, the verification efficiency is greatly improved, and the quality and the reliability of reading are improved.
Description
Technical Field
The invention relates to the technical field of digital automatic identification, in particular to a method and a device for identifying a reading image of an electronic water meter based on an improved threading method.
Background
The water meter is a main measuring device for water selling of water supply enterprises, the measuring accuracy is a key index for measuring the quality of the water meter, and the vital interests of the water supply enterprises and water consumers are directly influenced. The verification work of the water meter is a basic means for checking whether the metering accuracy of the water meter is qualified or not, and comprises 4 verification items, namely appearance, mark and seal, electronic device function, sealing performance and indication error. The indication error calibration project is mainly developed by adopting a comparison method, namely based on a continuity equation, the accumulated flow passing through the water meter is equal to the volume passing through the metering standard, and therefore the accuracy of the water meter is determined.
The water meters are classified into a mechanical water meter, a mechanical water meter with an electronic device and an electronic water meter according to the working principle and the structural characteristics. At present, most electronic water meter manufacturers and legal metering verification mechanisms verify the water meters in a manual operation or semi-automatic mode, the readings of the water meters are manually recorded, and the workload is large; the subjective error of the calibrating personnel is difficult to control, so that the metering accuracy is not high. In the accumulated flow reading link of the electronic water meter indication error verification project, the automatic detection mode is not ideal, and mainly carries out image recognition on the numbers of the liquid crystal display screen of the meter to be detected, and the image recognition has three modes: the identification method based on the neural network, the template matching identification method and the threading identification method.
The principle of electronic water meter reading image recognition based on the neural network is as follows: the behavior characteristics of an animal neural network are simulated, a distributed information processing mathematical model is constructed, and a large number of interrelations and parameters between nodes are set according to existing digital samples, so that unknown numbers are processed and analyzed. The neural network has the important characteristics of self-learning and self-adapting capability, analyzes related rules and containment relations between objects to be summarized according to a batch of input-output data which are provided in advance and correspond one by one, and can input new data according to the rules after analysis to finish result calculation, wherein the process is called training. Because a large amount of prior information (digital samples) is needed in the training process of the digital recognition model based on the neural network, the training process is complex, the recognition efficiency can be reduced, and certain requirements are imposed on operating equipment and computing resources.
The principle of the electronic water meter reading template matching identification method is as follows: extracting the numbers of the liquid crystal screen of the water meter, carrying out similarity calculation on the numbers to be recognized and the standard number templates one by one, measuring the difference between the numbers to be recognized and the characteristics of each number template, wherein the greater the similarity is, the smaller the difference is, and taking the corresponding standard number template with the maximum similarity as a final recognition result. The template matching method needs to establish a standard template library, has a large workload, requires the calculation of the similarity between the number to be identified and each standard number template, and has a large calculation amount and long verification time.
The principle of the electronic water meter reading threading identification method is as follows: as shown in fig. 1, seven segments a, b, c, d, e, f and g of the nixie tube are threaded. When 7 scan lines a, b, c, d, e, f, g have intersections with the number field, the intersection information is set to 1, and the non-intersection information is set to 0. The intersection information of the scanning lines a, b, c, d, e, f and g and each digital field is arranged in sequence and can be translated into a digital translation table, and the type of the number can be judged according to the translation table. The method is simple in principle, certain redundancy still exists in method design, the verification efficiency needs to be improved, the inclination condition of the nixie tube is not considered, and the verification accuracy is low.
In summary, the existing image recognition method for electronic water meter reading has disadvantages, the reading success rate and/or efficiency still need to be improved, and it is urgently needed to develop an image recognition method and device with accurate reading, high efficiency and strong anti-interference capability so as to realize efficient and accurate verification of the electronic water meter.
Patent documents CN108764234A, CN109858480A, CN110210477a and CN112686264a both disclose methods for identifying digital instruments, but the methods used are relatively complex, and the identification efficiency and accuracy are relatively low.
Disclosure of Invention
In order to solve at least one technical problem existing in the background art, the invention provides a method and a device for identifying a reading image of an electronic water meter based on an improved threading method.
In order to realize the purpose, the technical scheme of the invention is as follows:
in a first aspect, the invention provides an electronic water meter reading image recognition method based on an improved threading method, which comprises the following steps:
inputting electronic water meter reading images, adopting improved threading method identification to the water meter reading numbers in the images, the improved threading method identification comprises:
scanning from top to bottom from the midpoint of the upper boundary of the number, scanning from left to right from the 3/4 midpoint of the left boundary of the number and the 1/4 midpoint of the left boundary of the number respectively, and determining the number type according to the results of three times of scanning to obtain the accumulated flow indicating value of the detected water meter.
In a second aspect, the present invention provides an electronic water meter reading image recognition apparatus based on an improved threading method, including:
the visual detection module is used for shooting the liquid crystal display screen image of the detected water meter in real time to obtain an electronic water meter reading image;
the host computer for receive the electronic type water gauge reading image that visual detection module transmitted, adopt the improvement threading method discernment to the water gauge reading number in the image, improve threading method discernment and include:
scanning from top to bottom from the midpoint of the upper boundary of the number, scanning from left to right from the 3/4 midpoint of the left boundary of the number and the 1/4 midpoint of the left boundary of the number respectively, and determining the number type according to the results of three times of scanning to obtain the accumulated flow indicating value of the detected water meter.
Further, before the water meter reading number in the image is identified by adopting the improved threading method, the method further comprises the following steps:
and positioning and segmenting the nixie tubes, performing image affine transformation by using the top point of each nixie tube, and performing inclination correction on the numbers.
Further, after the inclination correction is performed on the numbers, the method further comprises the following steps: preprocessing a digital image, wherein the preprocessing of the digital image comprises image expansion processing, graying and binaryzation; after the digital image is preprocessed, each digit is identified by adopting an improved threading method.
Further, the visual detection module comprises a camera and an image processing unit; the camera shoots the image of the liquid crystal display screen of the detected water meter in real time and uploads the image to the upper computer; the image processing unit is used for marking the reference positions of four vertexes corresponding to each nixie tube in the liquid crystal display screen image of the water meter, detecting the actual positions of the vertexes, and feeding back the position deviation of the vertexes and the actual positions to the lower computer in real time;
the device further comprises:
and the positioning module is used for receiving a control instruction of the lower computer, comprises a pipe direction position adjusting unit and a radial position adjusting unit, and is used for adjusting the pipe direction and radial positions of the vision detection module, and the camera of the vision detection module vertically irradiates the front surface of the detected water meter to realize the coincidence of the vertex reference position and the actual position of the nixie tube.
Further, the apparatus further comprises:
the communication module is used for transmitting the reading image of the electronic water meter at the beginning of the meter reading period shot by the visual detection module to an upper computer so as to carry out digital image identification, and the identified and converted accumulated flow value is an initial water quantity settlement value; when the water fee settlement is carried out at the end of the charging period, the camera of the visual detection module shoots a dial picture again, the accumulated flow value obtained by identification and conversion is the accumulated flow value at the end of the user period, and the difference between the accumulated flow value at the end of the period and the initial value of the water amount settlement is the settlement water amount of the period;
all images and data are transmitted to a background server system of a water supply company and a user mobile terminal in a wireless remote mode through a communication module.
Further, the determining the number type according to the result of the three times of scanning to obtain the accumulated flow indication value of the detected water meter includes:
obtaining corresponding counting variable values count1, count2 and count3 by three times of scanning, wherein the count1, the count2 and the count3 are all assigned to be 0 under an initial condition; in the first scanning, when the pixel value of the first row is 0 or the pixel value of the previous row is 255 and the pixel value of the current row is 0, the count variable count1 is added with 1; in the second scanning, when the pixel value of the first column is 0, the count variable count2 is added with 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count2 is added with 10; in the third scanning, when the pixel value of the first column is 0, the count variable count3 is added with 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count3 is added with 10; according to the counting rule of the variables and the characteristics of the numbers from 0 to 9, the type of the numbers can be judged, and further the accumulated flow indication value of the water meter can be obtained.
Compared with the prior art, the invention has the beneficial effects that:
in the traditional threading method identification mode, 7 lines need to be scanned, so that the identification efficiency is relatively low; the improved threading method only needs to scan 3 lines, so that the image recognition algorithm is greatly simplified, the running time of the image recognition algorithm is shortened, the automatic reading efficiency of the water meter is improved, the verification efficiency is greatly improved, and the quality and the reliability of reading are improved.
Drawings
FIG. 1 is a schematic view of a conventional threading method for digital recognition;
fig. 2 is a flowchart of an electronic water meter reading image recognition method based on an improved threading method according to an embodiment of the present invention;
FIG. 3 is a schematic view of 3 threading scans of a nixie tube;
FIG. 4 is a schematic view of the nixie tube positioning and inclination correction;
fig. 5 is a schematic composition diagram of an electronic water meter reading image recognition device based on an improved threading method according to an embodiment of the present invention;
in the figure: 51. an upper computer; 52. a lower computer; 53. a positioning module; 54. a visual detection module; 55. a communication module; 56. a water supply company service system; 57. a user mobile terminal; 531. a pipe position adjusting unit; 532. a radial position adjusting unit; 541. a camera; 542. an image processing unit.
Detailed Description
The embodiment is as follows:
the technical solution of the present invention is further described below with reference to the accompanying drawings and examples.
Example 1:
at present, most electronic water meter manufacturers and detection devices of legal metering verification mechanisms verify water meters in a manual operation or semi-automatic mode, and technicians are mainly used for manually copying the accumulated flow of the detected water meters; the workload of technical personnel is large, the working efficiency is low, the reading error of human factors exists, and the verification quality is not easy to control.
In the link of detecting the accumulated flow of the detected meter of the indicating value error detection project, the automatic detection mode is not ideal. The existing automatic detection mode mainly carries out image recognition on the numbers of the liquid crystal display screen of the detected meter, and the problems of low detection efficiency exist.
For the electronic water meter reading identification method based on the neural network, a large amount of prior information (digital samples) is needed, the method is complex, the identification efficiency can be reduced, and certain requirements are imposed on operating equipment and operation resources; for the digital template matching identification method, a standard template library needs to be established, the workload is large, the similarity between the number to be identified and each standard digital template needs to be calculated, the calculation amount is large, and the identification time is long; for the traditional threading identification method, certain redundancy still exists in the method design, the identification efficiency needs to be improved, the condition that the nixie tube inclines is not considered, and the identification accuracy is low.
Therefore, the embodiment provides an electronic water meter reading image recognition method based on an improved threading method, as shown in fig. 2, including:
inputting electronic water meter reading images, adopting improved threading method identification to the water meter reading numbers in the images, the improved threading method identification comprises:
firstly, scanning from top to bottom from the midpoint of the upper boundary of the number, then respectively scanning from the left to right from the midpoint of 3/4 of the left boundary of the number and the midpoint of 1/4 of the left boundary, and determining the number type according to the results of three times of scanning to obtain the accumulated flow indicating value of the detected water meter.
As shown in fig. 1, in the conventional threading method, 7 lines (a-g) need to be scanned, and the recognition efficiency is relatively low; as shown in FIG. 3, the improved threading method only needs to scan 3 lines for identification, so that the image identification algorithm is greatly simplified, the running time of the image identification algorithm is shortened, the automatic reading efficiency of the water meter is improved, the verification efficiency is greatly improved, and the quality and the reliability of the reading are improved.
As an improvement of the electronic water meter reading image recognition method based on the improved threading method in this embodiment, before the water meter reading number in the image is recognized by the improved threading method, the method further includes:
and positioning and segmenting the nixie tubes, performing image affine transformation by using the top point of each nixie tube, and performing inclination correction on the numbers. Therefore, as shown in fig. 4, the digital correction of the inclination can be realized, the preparation is made for the subsequent verification work, and the improvement of the identification precision is facilitated. Compared with the traditional threading method, the processing method has the advantages that the identification precision is higher, and the error is smaller.
As another improvement of the electronic water meter reading image recognition method based on the improved threading method in the embodiment, after the inclination correction is performed on the numbers, the method further includes:
preprocessing a digital image, wherein the preprocessing of the digital image comprises image expansion processing, graying and binaryzation; after the digital image is preprocessed, each digit is identified by adopting an improved threading method. Through image expansion processing, the lines of the numbers become thicker, and the image recognition is easier; the collected three-channel image is simplified into a single-channel image through image graying processing; after the image binarization processing, the number becomes black, i.e., the pixel value is 0, and the remaining portion becomes white, i.e., the pixel value is 255. Then adopting improved threading method to make the above-mentioned processed digital image along the line segment
、
And
3 scans were performed: the scanning is performed from the top to the bottom from the middle point of the upper boundary of the number, and then from the left to the right from the middle point of 3/4 of the left boundary of the number and the middle point of 1/4 of the left boundary of the number respectively. Line segment
The small rectangular frame is divided into a left part and a right part, namely a line segment
And
the small rectangular frame is divided into an upper part, a middle part and a lower part. Line segment
、
And
corresponding to the count variables count1, count2 and count3, respectively, under the initial condition, count1, count2 and count3 are all assigned to be 0.
For line segment
That is, the pixels are scanned line by line from top to bottom at the midpoint of the upper boundary of the rectangular frame, after morphological preprocessing, the digital part is black, the pixel value is 0, the rest part is white, the pixel value is 255, and when the pixel value of the first line is 0 or the pixel value of the previous line is 255 and the pixel value of the current line is 0, the count variable count1 is incremented by 1. The partial numbers can be preliminarily determined according to the value of the count1 after scanning: when count1 is 0, the number is 1; when count1 is 1, the number is one of 4 and 7; when count1 is 2, the number is 0; when count1 is 3, the number is one of 2, 3, 5, 6, 8, 9. I.e. the numbers 0 and 1 can be matched in the first scan, the remaining 2, 3, 4, 5, 6, 7, 8, 9 need to be matched for decision in the following scans.
For line segment
That is, the pixels are scanned from left to right in 3/4 of the left border of the rectangular frame, when the pixel value of the first column is 0, the count variable count2 is incremented by 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count2 is incremented by 10. And combining the value of count1 and the value of count2 after the last scanning step to judge that: when count1 is 1 and count2 is 11, the number is 4; when count1 is 1 and count2 is 10, the number is 7; when count1 is 3, count2 is 1, the number is one of 5 and 6; when count1 is 3 and count2 is 10, the number is one of 2 and 3; when count1 is 3 and count2 is 11, the number is one of 8 and 9. I.e. the numbers 4 and 7 can be matched in the second scan, the remaining 2, 3, 5, 6, 8, 9 need to be matched for decision in the subsequent scans.
For line segment
That is, the pixels are scanned from left to right in 1/4 midpoint of the left boundary of the rectangular frame, when the pixel value of the first column is 0, the count variable count3 is added by 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count3 is added by 10. And combining the value of count2 and the value of count3 after the last scanning step to judge that: when count2 is 1 and count3 is 11, the number is 6; when count2 is 1 and count3 is 10, the number is 5; when count2 is 10 and count3 is 1, the number is 2; when count2 is 10 and count3 is 10, the number is 3; when count2 is 11 and count3 is 11, the number is 8; when count2 is 11 and count3 is 10, the number is 9.
According to the counting rule of the variables and the characteristics of the numbers from 0 to 9, different counting variable values can be obtained after 3 times of scanning, as shown in table 1.
TABLE 1 different numbers correspond to the number of counting variables
As can be seen from table 1, the combination of count1, count2 and count3 corresponding to each number is different, so that 10 numbers can be distinguished. Compared with the traditional threading method, the method only has 3 scanning lines, the running speed is higher, and the identification efficiency is greatly improved.
And finally, obtaining an accumulated flow indicating value of the detected water meter according to a digital identification result, calculating an indicating value error of the detected water meter by combining an actual volume measured by a standard device, and automatically generating an original record and a verification certificate, thereby realizing the automation of the whole verification process and improving the reading efficiency, the verification accuracy and the reliability of the water meter.
Example 2:
referring to fig. 5, the embodiment provides an electronic water meter reading image recognition device based on an improved threading method, and the device includes an upper computer 51, a lower computer 52, a positioning module 53, and a visual detection module 54.
The visual detection module 54 includes a camera 541 and an image processing unit 542, where the camera 541 is used to continuously shoot video at high speed and upload the video to the upper computer 51 (computer) to perform image recognition processing according to the steps of the method described in embodiment 1, that is, load a dial image, perform character positioning, segmentation, tilt correction (affine transformation of the image), image preprocessing (including image expansion, image graying, and binarization), image scanning (scanning from top to bottom from the midpoint of the upper boundary of the number, and then scanning from left to right from the midpoint of 3/4 of the left boundary and the midpoint of 1/4 of the left boundary of the number), and output a number recognition result. After the verification is started, the liquid crystal display screen images are continuously acquired, and the accumulated flow value of the detected meter is identified and converted, so that high-precision water flow automatic detection can be realized, subjective errors or errors of manual reading are avoided, the verification work is scientific, accurate and efficient, and the labor cost is reduced.
The image processing unit 542 is configured to mark reference positions of four vertices corresponding to each nixie tube in an image of a liquid crystal display of the water meter under test, detect actual positions of the vertices, and feed back a position deviation between the reference positions and the actual positions to the lower computer 52 in real time, in this embodiment, the lower computer 52 is a Programmable Logic Controller (PLC) to adjust a position of the visual detection module through the positioning module 53, the positioning module 53 is mounted with the visual detection module 54, which includes a tube direction position adjusting unit 531 (along a detection pipeline direction, referred to as a tube direction for short) and a radial position adjusting unit 532 (along a vertical pipeline direction, referred to as a radial direction for short) for fine adjustment of the tube direction and radial position of the visual detection module 54, and the camera vertically irradiates the front surface of the water meter under test. If the actual position of the top point of the nixie tube is at the tube direction deviation reference position, the left and right positions of the camera are adjusted by rotating a knob of the tube direction position adjusting unit 531; if the actual position of the top point of the nixie tube deviates from the reference position in the radial direction, the front and back positions of the camera are adjusted by rotating the knob of the radial position adjusting unit 532. Because the requirement on the detection precision is high, the fault-tolerant rate is between 1 and 2 mm, the visual detection module is required to shoot images in real time, the vertex position of each nixie tube is continuously detected, the vertex position is fed back to the lower computer 52 in real time according to the deviation between the vertex position and the vertex position, the rotation of the positioning module knob is adjusted, the superposition of the vertex reference position and the actual position of the nixie tube is finally realized, and the automatic reading efficiency of the water meter is improved.
When the water meter to be detected flows for a specified time or the water consumption is detected, the detection of the flow point is completed; after the verification of the three flow points of the minimum flow, the boundary flow and the common flow is finished, the verification is finished, and the fixed shaft of the visual detection module is controlled to ascend and then is retracted so that a verifier can conveniently detach the meter.
As an optimization of the electronic water meter reading image recognition device based on the improved threading method, the device further comprises a communication module 55, wherein the communication module 55 comprises a GPRS unit, a battery box and a light supplement lamp, and the device is installed on a water meter without a remote transmission function and can realize remote meter reading. The camera 541 shoots a dial plate photograph at the beginning of a meter reading period and transmits the dial plate photograph to the remote upper computer 51 for digital image recognition of the liquid crystal display screen, and the recognized and converted accumulated flow value is an initial water quantity settlement value; when the water fee settlement is carried out at the end of the charging period, the dial photograph is taken again, the accumulated flow value obtained by identification and conversion is the accumulated flow value at the end of the user period, and the difference between the accumulated flow value at the end of the period and the initial value of the water amount settlement is the settlement water amount of the period. All images and data are transmitted wirelessly through the communication module 55 and can be transmitted to the water supply company service system 56 and the user mobile terminal 57. Therefore, the water consumption value of the water meter can be obtained regularly without entering a home, a remote meter reading function is realized, a water supply company can conveniently master the water consumption of a user and carry out scientific and intelligent management on the water consumption condition, and the problems of difficulty in meter reading and large workload of entering the home are solved. The user can also check the water consumption condition of the user through the mobile terminal at any time and any place.
The above embodiments are only for illustrating the technical concept and features of the present invention, and the purpose thereof is to enable those skilled in the art to understand the contents of the present invention and implement the present invention accordingly, and not to limit the protection scope of the present invention accordingly. All equivalent changes or modifications made in accordance with the spirit of the present disclosure are intended to be covered by the scope of the present disclosure.
Claims (8)
1. An electronic water meter reading image recognition method based on an improved threading method is characterized by comprising the following steps:
inputting electronic water meter reading images, adopting improved threading method identification to the water meter reading numbers in the images, the improved threading method identification comprises:
scanning from top to bottom from the midpoint of the upper boundary of the number, scanning from left to right from the midpoint of 3/4 of the left boundary of the number and the midpoint of 1/4 of the left boundary of the number respectively, determining the number type according to the results of the three times of scanning, and obtaining the accumulated flow indication value of the detected water meter;
the step of determining the digital type according to the results of the three times of scanning to obtain the accumulated flow indication value of the detected water meter comprises the following steps:
obtaining corresponding counting variable values count1, count2 and count3 by three times of scanning, wherein the count1, the count2 and the count3 are all assigned to be 0 under an initial condition; in the first scanning, when the pixel value of the first row is 0 or the pixel value of the previous row is 255 and the pixel value of the current row is 0, the count variable count1 is added with 1; in the second scanning, when the pixel value of the first column is 0, the count variable count2 is added with 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count2 is added with 10; in the third scanning, when the pixel value of the first column is 0, the count variable count3 is added with 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count3 is added with 10; and judging the type of the number according to the counting rule of the variable and the characteristics of the numbers from 0 to 9 to obtain the accumulated flow indicating value of the water meter.
2. The method for recognizing the reading image of the electronic water meter based on the improved threading method as claimed in claim 1, further comprising before recognizing the reading number of the water meter in the image by the improved threading method:
and positioning and segmenting the nixie tubes, performing image affine transformation by using the top point of each nixie tube, and performing inclination correction on the numbers.
3. The method for recognizing the reading image of the electronic water meter based on the improved threading method as claimed in claim 2, further comprising, after the step of performing the inclination correction on the number: preprocessing a digital image, wherein the preprocessing of the digital image comprises image expansion processing, graying and binaryzation; after the digital image is preprocessed, each digit is identified by adopting an improved threading method.
4. The utility model provides an electronic type water gauge reading image recognition device based on improve threading method which characterized in that includes:
the visual detection module is used for shooting the image of the liquid crystal display screen of the detected water meter in real time to obtain a reading image of the electronic water meter;
the host computer for receive the electronic type water gauge reading image that visual detection module transmitted, adopt the improvement threading method discernment to the water gauge reading number in the image, improve threading method discernment and include:
scanning from top to bottom from the midpoint of the upper boundary of the number, scanning from left to right from the 3/4 midpoint of the left boundary of the number and the 1/4 midpoint of the left boundary of the number respectively, and determining the number type according to the results of three times of scanning to obtain the accumulated flow indicating value of the detected water meter;
the step of determining the digital type according to the results of the three times of scanning to obtain the accumulated flow indication value of the detected water meter comprises the following steps:
obtaining corresponding counting variable values count1, count2 and count3 by three times of scanning, wherein the count1, the count2 and the count3 are all assigned to be 0 under an initial condition; in the first scanning, when the pixel value of the first row is 0 or the pixel value of the previous row is 255 and the pixel value of the current row is 0, the count variable count1 is added with 1; in the second scanning, when the pixel value of the first column is 0, the count variable count2 is added with 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count2 is added with 10; in the third scanning, when the pixel value of the first column is 0, the count variable count3 is added with 1, and when the pixel value of the last column is 255 and the pixel value of the current column is 0, the count variable count3 is added with 10; and judging the type of the number according to the counting rule of the variable and the characteristics of the numbers from 0 to 9 so as to obtain the accumulated flow indicating value of the water meter.
5. The electronic water meter reading image recognition device based on the improved threading method as claimed in claim 4, wherein:
before the water meter reading number in the image is identified by adopting an improved threading method, the method further comprises the following steps:
and positioning and segmenting the nixie tubes, performing image affine transformation by using the top point of each nixie tube, and performing inclination correction on the numbers.
6. The electronic water meter reading image recognition device based on the improved threading method as claimed in claim 5, wherein: after the inclination correction is carried out on the digital image, the method also comprises the following steps: preprocessing a digital image, wherein the preprocessing of the digital image comprises image expansion processing, graying and binaryzation; after the digital image is preprocessed, each digit is identified by adopting an improved threading method.
7. The electronic water meter reading image recognition device based on the improved threading method as claimed in claim 5,
the visual detection module comprises a camera and an image processing unit; the camera shoots the image of the liquid crystal display screen of the detected water meter in real time and uploads the image to the upper computer; the image processing unit is used for marking reference positions of four vertexes corresponding to each nixie tube in the detected water meter liquid crystal display screen image, detecting actual positions of the vertexes and feeding back position deviation of the vertexes to the lower computer in real time;
the device further comprises:
and the positioning module is used for receiving a control instruction of the lower computer, comprises a pipe direction position adjusting unit and a radial position adjusting unit, and is used for adjusting the pipe direction and radial positions of the vision detection module, and the camera of the vision detection module vertically irradiates the front surface of the detected water meter to realize the coincidence of the vertex reference position and the actual position of the nixie tube.
8. The electronic water meter reading image recognition device based on the improved threading method as claimed in claim 4 or 7, wherein the device further comprises:
the communication module is used for transmitting the reading image of the electronic water meter to a remote upper computer at the beginning of a meter reading period shot by the visual detection module so as to carry out digital image recognition, and the recognized and converted accumulated flow value is an initial water amount settlement value; when the water fee settlement is carried out at the end of the charging period, the camera of the visual detection module shoots a dial picture again, the accumulated flow value obtained by identification and conversion is the accumulated flow value at the end of the user period, and the difference between the accumulated flow value at the end of the period and the initial value of the water amount settlement is the settlement water amount of the period;
all images and data are transmitted wirelessly through the communication module and transmitted to the water supply company service system and the user mobile terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210861687.2A CN114926841B (en) | 2022-07-22 | 2022-07-22 | Electronic water meter reading image identification method and device based on improved threading method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210861687.2A CN114926841B (en) | 2022-07-22 | 2022-07-22 | Electronic water meter reading image identification method and device based on improved threading method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114926841A CN114926841A (en) | 2022-08-19 |
| CN114926841B true CN114926841B (en) | 2022-11-08 |
Family
ID=82815753
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210861687.2A Active CN114926841B (en) | 2022-07-22 | 2022-07-22 | Electronic water meter reading image identification method and device based on improved threading method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114926841B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117011167B (en) * | 2023-07-03 | 2024-08-20 | 广东盈科电子有限公司 | Image positioning correction method, system and device for nixie tube and storage medium |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106709484B (en) * | 2015-11-13 | 2022-02-22 | 国网吉林省电力有限公司检修公司 | Digital identification method of digital instrument |
| CN108491844A (en) * | 2018-02-07 | 2018-09-04 | 西安工程大学 | Water meter automatic checkout system based on image procossing and its image processing method |
| CN108549889B (en) * | 2018-03-14 | 2021-07-27 | 南京航空航天大学 | Simple print form number identification method |
| CN109255344B (en) * | 2018-08-15 | 2022-02-18 | 华中科技大学 | Machine vision-based digital display type instrument positioning and reading identification method |
| CN109753476A (en) * | 2019-01-11 | 2019-05-14 | 北京师范大学 | A kind of bill scans financial special calculator and application method |
| EP3772015B1 (en) * | 2019-07-31 | 2023-11-08 | MyScript | Text line extraction |
-
2022
- 2022-07-22 CN CN202210861687.2A patent/CN114926841B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114926841A (en) | 2022-08-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Deng et al. | Automatic indoor construction process monitoring for tiles based on BIM and computer vision | |
| CN108124487B (en) | Cloud meter reading method and device | |
| CN112906694B (en) | Reading correction system and method for transformer substation inclined pointer instrument image | |
| CN110119680B (en) | Automatic error checking system of regulator cubicle wiring based on image recognition | |
| CN105091922B (en) | A kind of pointer gauge Recognition of Reading method based on virtual dial plate | |
| CN105303168A (en) | Multi-view pointer type instrument identification method and device | |
| WO2016112669A1 (en) | Character information recognition method based on image processing | |
| CN107507174A (en) | Power plant's instrument equipment drawing based on hand-held intelligent inspection is as recognition methods and system | |
| CN101324452A (en) | Method for automatically detecting pointer instrument | |
| CN112801094B (en) | Pointer instrument image inclination correction method | |
| CN110146019A (en) | A kind of workpiece size on-line checking and automatic compensating method based on industrial camera | |
| CN111368906A (en) | Pointer type oil level indicator reading identification method based on deep learning | |
| CN114926841B (en) | Electronic water meter reading image identification method and device based on improved threading method | |
| CN116465315B (en) | Automatic screen quality detection method and system | |
| CN113537075A (en) | Pointer instrument reading identification method and device, storage medium and electronic equipment | |
| CN114663744A (en) | Instrument automatic identification method and system based on machine learning | |
| CN116434266A (en) | Automatic extraction and analysis method for data information of medical examination list | |
| CN113034620A (en) | Correction method, correction device, computer-readable storage medium, and computer apparatus | |
| CN116612180B (en) | Commodity quantity detecting system capable of being maintained in real time | |
| CN114092542B (en) | Bolt measurement method and system based on two-dimensional vision | |
| CN109410210A (en) | Bar code printing quality detection method based on machine vision | |
| CN115546539A (en) | Magnetic flip plate liquid level reading method and device based on machine vision and readable medium | |
| CN118408478B (en) | Accurate measuring device of wisdom construction based on BIM technique | |
| CN117392656A (en) | Liquid crystal and character wheel double-dial accurate reading method integrating deep learning and traditional image processing | |
| CN114782436B (en) | PCB defect detection method, circuit board metering measurement method and manufacturing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |