KR102322066B1 - Method for gas meter management using image analysis technology - Google Patents

Abstract

The present invention relates to a method for managing a gas meter. Specifically, the present invention relates to the method that can accurately and conveniently perform a meter reading work, a meter replacement work, a meter management work when moving out, a collection work, and the like by an external meter reading person using an image analysis technology. The method comprises: a whether or not registered terminal checking step; a meter reading data transmitting step; an image generating step; a meter reading meter image transmitting step; a transmitting step to a control server; a household number checking step; an updating step; and a calculating step.

Description

Method for gas meter management using image analysis technology

The present invention relates to a gas meter management method, and more specifically, to a method that can accurately and conveniently perform meter reading work by an external meter reading source, meter replacement work, meter management work when moving out, collection work, etc. using image analysis technology it's about

In the conventional gas meter meter reading, an employee of a gas supplier or a meter reader who has been trained to read the guideline value from the gas supplier visits each household and checks the meter one by one, then checks the instrument number and guideline value written on the meter separately. It was performed by handwriting on a record-only terminal such as a notebook or PDA. There are several types of meters, but the location where the object number is written on each meter, the method of reading the object number, and the method of reading the guide value are all different, so skilled personnel are required.

Moreover, it is important to determine the order of the meter reading, as one meter reader has to meter many generations of gas meters within a limited time. However, due to the different bill payment schedules for each household, the meter reading schedule is different (for example, the meter reading schedule is different for a household that pays on the 1st of every month and a household that pays on the 15th of every month), the location of the household is different in each region, and the Since the locations are all different, it is difficult to determine an optimized meter reading sequence. Conventionally, the order of the meter reading was determined by the experience of the meter reader, so it took a lot of time for a beginner with little experience to read a household of the same size.

Accordingly, a patent for performing meter reading using artificial intelligence has been proposed. Korean Patent Registration No. 10-2162641 proposes a method for recognizing meter information based on deep learning. Here, when the meter reader acquires the image of the meter to be read, the object number and the guide value are automatically checked. The problem that only an experienced meter reader should read the gas meter has been solved to some extent by the present prior art.

However, even when the present prior art is applied, many problems are still not solved.

First, the results of artificial intelligence are uncertain, causing a fatal problem of incorrect billing. If you have been billed for gas from another household, there is a lot of difficulty in resolving it if you have even applied for automatic payment. In fact, the same problem occurred when the article number was incorrectly written in the conventional meter reading method written by hand, and it is expected that the frequency of this problem will increase in the present prior art. This is because gas meter manufacturers are diversifying, and even one manufacturer is selling many types of gas meters. Even with the development of artificial intelligence, since the accuracy of identification of the object number by image interpretation is not 100%, the problem of incorrect billing is still not solved.

Second, the meter reading sequence still depends on experience. Even if this prior art is applied, a meter reading officer must first visit the household to take a picture of the meter. As a result, as in the conventional method, since the meter reading source must move according to the meter reading sequence, an optimized meter reading sequence cannot be set unless there is still a lot of experience, and it takes a lot of time to read multiple generations.

Third, a large number of meter reading sources are required at the gas supplier. As described above, a meter reader must visit the household, and an inexperienced meter reader cannot optimize the meter reading sequence, so many meter readers are still needed. That is, with this prior art alone, the gas supplier must still educate the meter reading personnel and secure many meter reading personnel.

Therefore, the gas supplier does not need to secure many meter reading sources, and the demand for a technology capable of optimizing the meter reading sequence and having high meter reading accuracy regardless of experience is, like this prior art, artificial intelligence-based meter reading technology.

Even in this developed state, it is still present.

On the other hand, apart from the above technology, a lot of manpower and cumbersome steps are still required in the various tasks performed at the customer center of the gas supplier.

For example, when a meter needs to be replaced due to malfunction or old age, a gas supplier employee visits the household to remove the existing meter and put it in the ENG (warehouse) while installing a new meter. The guide value of the removed meter must be separately checked and notified to the customer center, and the customer center must calculate the fee based on this so that it can be charged for the next billing. In addition, it is necessary to check that the guide value of the newly installed meter is 0, and even if it is not 0, this must also be reported to the customer center so that the correct rate can be calculated at the next meter reading. Therefore, when replacing the meter, it is necessary to check and inform the guide value several times, and this process cannot be done immediately, so there is a time difference of several days.

For another example, if a household gas user wants to move out due to moving, etc., the moving-out person must visit the customer center a few days before moving out or apply for moving out through the website or landline phone. Then, on the day of moving out, an available employee of the gas supplier visits the household directly to check the guide value and informs the customer center, which determines the fee based on this and informs the person moving out. The processing operation is terminated. This process has a time difference and is complicated and cumbersome. If the user directly checks the guideline value, the employee's household visit step can be omitted.

For another example, a separate collection agency, rather than the gas supplier, conducts collection activities for non-payment of gas bills. At this time, the collector must visit the household to determine whether collection is possible. do. However, even if the gas supply is stopped, the guideline value should be checked up to that point. Therefore, when the collection agency notifies the customer center that collection is impossible, the customer center staff separately visits the household again to check the guide value, and reflects this to determine the final payment fee. Two or more people visit the household, and the process is cumbersome and complicated.

(Patent Document 1) KR10-2162641B

(Patent Document 2) JP2020-027531A

(Patent Document 3) KR10-1897923B

The present invention seeks to solve the above-mentioned problems. It is intended to present a new gas meter meter reading management technology that utilizes the prior art of reading the object number and guide value using the meter image analysis technology, but can solve other problems that cannot be solved only with the prior art.

When using the meter image analysis technology, even if the recognition result of artificial intelligence is uncertain, we would like to propose a method with a double or triple confirmation system so that the generation indicated by the meter can be accurately identified.

In addition, we would like to propose a meter reading method that can be performed by an external meter reader, not an employee of the gas supplier. All procedures can be completed only by updating the image, and an optimized meter reading sequence can be automatically suggested, eliminating the need for external meter reading personnel to receive much training, and at least experience ensures accuracy and speed, saving manpower from the gas supplier’s point of view We would like to suggest an easy way to secure it.

In addition, we intend to propose a more effective management technique by optimizing and applying the meter image analysis technology to other technologies related to gas meter reading.

For example, we would like to propose a method in which image analysis technology can be applied to meter replacement, meter management when moving out, and collection work. Rather than simply applying image analysis technology, we want to introduce a more accurate authentication method to offset the shortcomings of the technology, and propose an optimized procedure for each management method.

In order to solve the above problems, (a1) the control server 300, the step of checking whether the request to access the terminal 100 is a registered terminal; (b1) transmitting the meter reading data of the meter reading target to the terminal 100; - The meter reading data includes a household number and an object number mapped thereto - (c1) generating a meter reading meter image in the terminal 100; (d1) transmitting, by the terminal 100, an image of the meter reading meter to the image analysis AI server 200; (e1) transmitting, by the image analysis AI server 200, to the control server 300 by using the meter reading meter image transmitted in the step (d1) to check the article number and the guide value; (f1) confirming the household number by using the household number confirmation unit 313 of the control server 300 using the confirmed unit number; (g1) the meter-reading household management unit 312 updates a guideline value using the identified household number and article number in the meter-reading data, updates the meter-reading meter image, and time information at which the meter-reading meter image is generated Checking to update the meter reading date and time; and (h1) calculating, by the rate calculating unit 320, the gas rate using the meter reading data and the guide value.

In addition, the meter reading meter image further includes GPS information of the terminal 100 at the time of image creation, and the meter reading data further includes an address mapped in advance to a household number, and the step (f1) includes: The number check unit 313 compares the GPS information included in the meter-reading meter image with the address previously stored in the meter-reading data as corresponding to the household number confirmed in step (f1). It preferably includes a step of verifying correctness.

In addition, the meter reading sequence suggestion unit 314 is configured to include, among the household numbers included in the meter reading data, a distance between an address of one household number and an address of another household number, a meter reading date and time input for any one household number, and By performing machine learning using the difference in meter reading date and time input for other household numbers and the order in which the meter reading was made by each household number as input data, machine learning is performed. After the step (g1), (i11) the meter reading sequence calculating unit 314 calculates the generation number confirmed in the step (f1). setting as a reference; (i12) The meter-reading sequence calculating unit 314 calculates the meter-reading sequence score for all other household numbers for which the meter-reading meter image is not updated, using the criteria, and identifies the household number with the highest meter-reading sequence score and ranks it granting; (i13) The meter reading sequence calculation unit 314 continuously repeats the step (i12) based on the ranked household number as a new standard, so that all other household numbers for which the meter reading meter image is not updated are ranked. and calculating a meter reading order using the assigned rank; and (i14) outputting, by the meter reading sequence calculating unit 314, the meter reading sequence on a map.

In addition, before the step (a1), (a01) registering the identifier of the terminal 100 in the external terminal management unit 311; (a02) assigning, by the external terminal management unit 311, a reference address to the identifier of the registered terminal 100; and (a03) allocating, by the external terminal management unit 311, an identifier of the terminal 100 to a household number having an address within a predetermined distance based on the reference address, the step (b1) Preferably, the meter reading data to be transmitted to the terminal 100 is data on the household number assigned in step (a03) with respect to the identifier of the terminal 100 .

In addition, (a04) the terminal 100 requesting the meter reading data of the object to be read; (a05) confirming, by the external terminal management unit 311, the identifier of the terminal 100 requested in the step (a04); (a06) transmitting, by the meter-reading generation management unit 312, meter-reading data for the generation number to which the identifier of the terminal 100 is assigned to the terminal 100; and (a07) outputting the meter reading data transmitted by the terminal 100, wherein the household information output in the step (a07) includes a household number, an article number, a meter reading date and time, a guide value, an address, a meter reading source, It includes a meter reading order and an image link, wherein the meter reading date and time is the most recent input as the meter reading date and time input in step (g1), and the meter reading source is assigned to each identifier of the terminal 100 confirmed in step (a05). and the meter reading order is the order confirmed in step (i12). When the address included in the household information is activated, the address is output along with the map, and when the image link included in the household information is activated, the (d1) ), it is preferable that the meter-reading image transmitted in step is output.

Another embodiment of the present invention for solving the above problems includes the steps of: (b21) transmitting the household number of the household to be replaced by the meter and the address assigned thereto from the control server 300 to the terminal 100; (c21) generating a demolition meter image in the terminal 100; (d21) the terminal 100 transmitting the image of the demolishing meter to the image analysis AI server 200; (e21) the image analysis AI server 200, using the demolition meter image transmitted in the step (d21) to check the object number and the guide value and transmit it to the control server 300; (f21) confirming the household number by using the household number confirmation unit 313 of the control server 300 using the confirmed unit number; (g21) the meter replacement management unit 315 determines whether the pre-input number of the demolition target meter is the same as the piece number confirmed in step (e21), and the pre-entered household number of the meter replacement target household Checking whether the household numbers confirmed in step (f21) are the same, and transmitting a demolition meter confirmation signal to the terminal 100 if they are all the same; and (h21) the meter replacement management unit 315 removes the guideline value by using the household number confirmed in the confirmed step (f21) and the appliance number confirmed in the step (e21) in the replacement meter data Preferably, the method includes updating the demolition date and time by updating the guide value, updating the demolition meter image, and confirming the time information at which the demolition meter image was generated.

In addition, after the step (h21), (c22) generating an installed meter image in the terminal 100; (d22) transmitting, by the terminal 100, an installed meter image to the image analysis AI server 200; (e22) the image analysis AI server 200, using the installed meter image transmitted in the step (d22) to check the equipment number and the guide value and transmit it to the control server 300; (g22) The meter replacement management unit 315 checks whether the equipment number of the meter to be installed in advance and the equipment number confirmed in step (e22) are the same, and if they are the same, the replacement confirmation is sent to the terminal 100 transmitting a signal; and (h22) the meter replacement management unit 315 uses, in the replacement meter data, the household number confirmed in step (f22) and the unit number confirmed in step (e22), (e22) Preferably, the method further includes updating the guideline value identified in the step to the installation guideline value, updating the installed meter image, and checking information about the time when the installed meter image was generated to update the installation date and time.

In addition, the demolition meter image further includes GPS information of the terminal 100 at the time of image creation, and in the step (f21), the household number confirmation unit 313 includes the GPS included in the demolition meter image. It is preferable to include the step of verifying the accuracy of the identified household number by comparing the information with the address transmitted in the step (b21).

In addition, in the step (h22), the meter replacement management unit 315 updates the guideline value using the household number confirmed in the step (f21) and the appliance number confirmed in the step (e22) in the meter reading data. and updating the replacement meter image, and updating the meter replacement date and time confirmed in step (h22) to the meter reading date and time.

Another embodiment of the present invention for solving the above problems, (a3) the control server 300, the step of checking whether the request to access the terminal 100 is a registered terminal; (b3) transmitting a customer identifier and a moving-out processing signal from the terminal 100 to the moving-out household management unit 316; (c3) generating a billing meter image in the terminal 100; (d3) transmitting, by the terminal 100, an image of the transfer meter to the image analysis AI server 200; (e3) the image analysis AI server 200, using the transfer meter image transmitted in the step (d3) to check the equipment number and guide value and transmit it to the control server 300; (f3) confirming the household number by the household number confirmation unit 313 of the control server 300 using the confirmed object number; (g3) the moving-out household management unit 316 checks whether the household number confirmed in the step (f3) is the same as the household number corresponding to the customer identifier transmitted in the step (b3), and if the same, the meter reading data , updating a guideline value by using the identified household number and appliance number, updating the moving-out meter image, and checking the time information at which the moving-out meter image was generated to update the meter reading date and time; (h3) calculating, by the fee calculating unit 320, a fee using the checked guide value and transmitting the calculation to the terminal 100; and (i3) generating, by the gas supply control unit 330, a signal for stopping the supply of gas transmitted to the identified household number.

In addition, the moving-out meter image further includes GPS information of the terminal 100 at the time of image creation, and the meter reading data further includes an address previously mapped to the household number, and in step (f3), the household number verifying, by the verification unit 313, the GPS information included in the moving-out meter image and the address previously stored in the meter reading data as corresponding to the identified household number, verifying the accuracy of the confirmed household number; It is preferable to include

Another embodiment of the present invention for solving the above problem, (a4) the control server 300, the step of checking whether the request to access the terminal 100 is a registered terminal; (b4) transmitting collection data including meter reading data of a predetermined generation as a collection target to the terminal 100; - The collection data includes a household number and an object number mapped thereto - (c4) generating a collection meter image in the terminal 100; (d4) transmitting, by the terminal 100, a collection meter image to the image analysis AI server 200; (e4) the image analysis AI server 200, using the collection meter image transmitted in the step (d1) to check the article number and the guide value and transmitting to the control server (300); (f4) confirming the household number by the household number confirmation unit 313 of the control server 300 using the confirmed unit number; (g4) the collection management unit 317 checks whether the household number confirmed in step (f4) is the same as the household number transmitted in step (b4), and if the same, in the collection data, the identified household Updating the guideline value using the number and the article number, updating the collection meter image, and checking the time information at which the collection meter image was generated to update the supply-oriented date and time; (h4) calculating, by the rate calculating unit 320, a rate using the checked guide value; and (i4) generating, by the gas supply control unit 330, a signal for stopping the supply of gas transmitted to the identified household number.

In addition, the collection meter image further includes GPS information of the terminal 100 at the time of image creation, the collection data further includes an address mapped in advance to a household number, and the step (f4) is, the household Verification of the accuracy of the identified household number by comparing, by the number verification unit 313, the address stored in advance in the collection data to correspond to the identified household number with the GPS information included in the collection meter image It is preferable to include

The following effects are achieved by the present invention.

First, the number of specialized meter reading personnel can be greatly reduced. It does not require much training, experience is not important, and the generation can be read with high accuracy by unskilled workers rather than professional meter readers. For this reason, an external meter reading source may be used, and in the case of an apartment, a management office manpower may be used, and depending on the user, the meter reading is also possible. Even without much experience, the meter reading route is proposed according to the present invention, so that the meter reader can move to the optimized route.

Second, the meter reading accuracy is greatly improved. When using image analysis technology, the recognition result of artificial intelligence is uncertain and the wrong number is recognized, causing a problem of incorrect billing. This problem is solved by verifying , household number, etc.

Third, when replacing the meter, the staff only needs to photograph the demolished meter and the installed meter. There is no need to ask the customer center for payment, and many procedures are omitted.

Fourth, the work of the moving-out person (customer) is simplified, and many steps of moving-out-related work processing are omitted. Accordingly, the moving-out does not need to apply for moving-out processing several days prior to moving out. The accuracy is high when the moving-in person directly checks the guide value.

Fifth, many steps are omitted even when the collection agency conducts collection activities. There is no need to ask the customer center to stop delinquency for households in arrears, and there is no need for the gas supplier's staff to visit the household to read the meter again. All procedures are completed only with the judgment of the collection officer and the photographing of the collection officer's image.

1 and 2 are schematic diagrams of a system for carrying out the method according to the invention, and FIG. 2 shows the control server of FIG. 1 in detail.
3 is a photograph of a meter for explaining the image analysis AI technology according to the present invention.
4 is a map for explaining the technique of the meter reading sequence according to the present invention.
5 to 7 are a format of data used in the method according to the present invention, Figure 5 explains the format of the meter reading data, Figure 6 describes the format of the replacement meter data, Figure 7 is the format of the collection data Explain.
8 to 11 are flow charts of the method according to the present invention, wherein FIG. 8 is a flow chart for explaining the meter reading task by an external meter reading source in Example 1, and FIG. 9 is a flowchart illustrating the meter replacement task in Example 2, 10 is a flowchart illustrating the meter management task when moving out of the third embodiment, and FIG. 11 is a flowchart illustrating the collection task of the fourth embodiment.

Hereinafter, "gas supplier" means a company that supplies and manages gas.

Hereinafter, "external meter reading personnel" refers to personnel belonging to the gas supplier, who directly visit each household and read the meter. Conventionally, gas meter meter reading was performed only by experienced gas supplier employees who received relevant training, but according to the present invention, anyone other than gas supplier employees may be an external meter reading source. It may be a form in which a separate operator other than the gas supplier establishes a meter reading company and hires employees, and it is also possible for apartment management office employees, community center employees, and users themselves. This is because, according to the present invention, the meter reading task is simplified enough to suffice to take only a photo of the meter, so that anyone can read the meter without professional training.

Hereinafter, "user" means a household resident who receives and uses gas.

Hereinafter, the term "collector" refers to an employee of a collection agency, which is a professional company for collecting gas bills in arrears, and is not an employee of a gas supplier, but refers to a person who directly visits the household to check whether collection is possible.

Hereinafter, the "meter image" means an image of the meter as shown in FIG. 3 . Below, depending on the purpose of measuring the image of the meter, it is divided into terms such as meter reading meter image, demolition meter image, installed meter image, move out meter image, and collection meter image.

An object of the present invention is not only to simplify a simple meter reading operation using a meter image analysis technology, but also to accurately and conveniently perform various tasks performed by a gas supplier by applying the technology.

We propose a new method for the following four tasks.

Example 1 - Meter reading work by an external meter reading source

Example 2 - Meter replacement work

Example 3 - Meter management work when moving out

Example 4 - Collection work

Hereinafter, each embodiment will be described in detail with reference to the drawings.

1. Description of the system

A system for carrying out the method according to the present invention will be described with reference to FIGS. 1 and 2 .

The system consists of a terminal 100 , an image analysis AI server 200 , and a control server 300 .

The terminal 100 may be a terminal of an external meter reading source (Embodiment 1), may be a terminal of a gas supplier employee (Embodiment 2), may be a user's terminal (Embodiment 3), or a terminal of a collection officer (implementation) Yes 4).

The terminal 100 may be any terminal having an information processing function. It may be anything such as a smartphone, tablet PC, or PDA. However, in order to carry out the present invention, it is preferable to include a camera capable of taking an image, it is more preferable to be able to record the time of taking an image, and it is more preferable to be able to specify a location for taking an image by having a GPS function. .

The image analysis AI server 200 receives the meter image photographed by the terminal 100 and functions to check the article number and the guide value. To this end, in conjunction with the object image database 290, in which various types of object images, object numbers and guide values written in different positions for each type of object image are databased, object numbers and instructions when checking the meter image The value can be checked by artificial intelligence.

3 shows one embodiment of a meter image. The position, size, number of numbers, etc. of the object number (upper box) and guide value (lower box) are all different for each meter. In the object image database 290, all images of currently commercialized meters are stored, and the location, size, number of numbers, etc. of the object number and the guide value are stored. The image analysis AI server 200 uses this to automatically classify what kind of meter it is when it obtains a meter image, uses this to check the location of the object number and guide value, and then uses the text reading technique to check the specific number By doing so, the object number and the guide value can be identified.

Other than that, any technology that checks the object number and guide value in the meter image can be applied. Since more specific details are prior art, detailed descriptions thereof will be omitted.

The control server 300 is a server operated by a gas supplier for the purpose of customer support such as meter reading. For example, it may be a server operated by a customer center of a gas supplier.

The control server 300 includes an external terminal management unit 311 , a meter reading household management unit 312 , a household number verification unit 313 , a meter reading sequence calculation unit 314 , a meter replacement management unit 315 , a moving out household management unit 316 , and a fee It includes a management unit 320 and a gas supply control unit 330 .

The external terminal manager 311 manages the terminal 100 accessing the control server 300 . In the case of an external terminal 100 other than a terminal directly connected to the control server 300 by wire or wirelessly, an application for registration is made through the external terminal management unit 311, and whether it is a registered terminal is checked.

The meter reading household management unit 312 checks a household that is to be periodically metered by receiving gas. To this end, it cooperates with the generation database 391 in which the meter reading data is stored and managed.

The household number confirmation unit 313 performs a function of confirming which household meter it is, when the equipment number is checked. To this end, it cooperates with the generation database 391 in which the meter reading data is stored and managed.

The meter reading sequence calculating unit 314 performs a function of calculating and providing an optimal sequence by calculating the meter reading sequence of the meter reading source. In order to calculate the optimal sequence, a machine-learning artificial intelligence model is used with data stored in the meter reading sequence database 392 .

4 shows the calculated flow of the meter reading. The number is the number of generations. The meter reading order calculating unit 314 according to the present invention may calculate a priority for each generation, and may then suggest a path through which the meter reading member should move on the map, that is, the meter reading order. Specific details will be described later.

The meter replacement management unit 315 manages information on a meter to be removed and a meter to be installed when the meter is replaced due to a regular replacement cycle, failure of the meter, aging, or the like. To this end, it cooperates with a meter database 393 in which replacement meter data including information on demolished meters and installed meters is stored and managed.

The moving-out household management unit 316 manages the household that has applied for moving out for reasons such as moving.

The collection management unit 317 performs management of households that become collections for reasons such as non-payment of fees. To this end, it cooperates with the collection database 394 in which collection data including information about the collection generation is stored and managed.

The rate calculation unit 320 performs a function of calculating a gas rate in a preset manner by using the recent guideline value and the current guideline value.

The gas supply control unit 330 generates a signal related to gas supply, such as stopping or resuming gas supply to a specific household. The generated signal is transmitted to the gas supplier server 400 . The gas supplier server 400 receives a signal from the gas supply control unit 330 to perform, manage, and record various tasks directly related to gas supply, such as stopping or resuming gas supply to a specific household.

2. Description of data

The data used in the present invention are meter reading data, replacement meter data, and collection data. It will be described with reference to FIGS. 5 to 8 .

The meter reading data is shown in FIG. 5 and may be stored in the generation database 391 .

The meter reading data can be checked, modified, deleted, etc. in the control server 300 , and can also be checked in the terminal 100 . However, it is sufficient to upload only the meter image in the terminal 100, and there is no need to directly input a specific object number, guide value, meter reading date, and the like. This is because, as will be described later, it is automatically written when the image is updated.

“Household number” is a number that identifies each household, and “equipment number” is a number that identifies each meter. In general, one unit number is assigned one-to-one to one household number.

"Address" is the address of that household. When the terminal 100 or the control server 300 activates, such as by clicking on the address, the address may be output together with a map. This will help the movement of the external meter reading source.

"Meter reading date and time" indicates the year, month, day, hour, and minute of the most recent meter reading. There is no need for an external meter reading person to directly input the meter reading date and time. As will be described later, the external meter reading source updates the image taken through the terminal 100. At this time, since the time point at which the image is generated is mapped to the image data, the time point is automatically registered as the meter reading date and time.

"Guideline value" is the guide value of the meter reading date and time. Similarly, the guide value checked by the image analysis AI server 200 is automatically updated, rather than directly input by an external meter reading source.

In other words, when an external meter reader updates the image, the object number is checked through image analysis to determine which generation the meter reading result is, and then the meter reading date and time checked here and the guide value are automatically input.

Meanwhile, the meter reading date and time and the guide value may be set so that only the most recent meter reading result is displayed, but the previous meter reading result is also stored in the household database 391 as a separate database.

"Meter reading officer" is the name of a person assigned to read the meter of the corresponding household, and may be an external meter reading source. The external meter reading source may register its own terminal 100 to the control server 300 through the external terminal management unit 311. At this time, the corresponding terminal identifier is registered and may be stored in the meter reading data.

"Meter-reading order" means the meter-reading priority. 5 , it can be seen that the meter reading was performed on December 1 in the first generation, but the meter reading was performed in the second generation on November 1st. If the current time is December 1, 2020, the first generation will not display anything in the meter reading sequence because the December meter reading has already been made. However, since the second generation has not yet been inspected in December, its priority will appear in the order of reading. In the embodiment shown in FIG. 5 , the priority of the second generation is “1”, and therefore, it is recommended that the meter reading officer “Gil-Dong Hong” go to the second generation to read the meter. The priorities appearing in the meter reading sequence are calculated by the meter reading sequence calculating unit 314, and details will be described later.

In the "image" column, when an external meter-reader uploads an image, a link display is displayed, and when the link is activated by the control server 300 or the terminal 100, the uploaded image is output. As shown in FIG. 5 , a link appears in the image column as the first generation has been read in December, but the link does not appear in the image column because the second generation has not yet been inspected in December.

Replacement meter data is shown in FIG. 6 and may be stored in meter database 393 .

Similarly, replacement meter data can be checked, modified, deleted, etc. in the control server 300 and can also be checked in the terminal 100, but it is sufficient to upload only the meter image in the terminal 100, and a specific piece number, guide value, etc. It is not necessary to directly input the meter reading date and time.

The replacement meter data is data about the operation of demolishing the currently installed meter and installing another meter, so the item number, guide value, image, etc. are separated separately for each demolition/installation. In addition, the time when the demolition meter image is generated becomes the demolition date and time, and the time when the installation meter image is generated becomes the installation date and time.

Collection data is shown in FIG. 7 and may be stored in a collection database 394 .

Similarly, the collection data can be checked, corrected, deleted, etc. in the control server 300, and can be checked in the terminal 100, but in the terminal 100, it is sufficient to upload only the meter image, and a specific article number, guide value, There is no need to directly input the meter reading date and time.

"Collection Officer" is the name of the person assigned to the household's collection task, and is an outside person who is not a gas supplier employee. You can register your terminal 100 to the control server 300 through the external terminal management unit 311, at this time, the terminal identifier is registered and can be stored in the collection data.

"Supply stop date and time" means the time when the gas supply to the household is stopped during the collection process, and is automatically entered as the time when the meter image is created.

3. Description of the method

In order to carry out the method according to the present invention, the image is taken by an external person who is not an employee of the gas supplier. Specifically, in Example 1, an external meter reading source, in Example 3, a user, and in Example 4, a collection officer must directly photograph the meter and connect to the control server 300 . To this end, first, the external terminal 100 must be registered with the control server 300 and authenticated.

The process will be described using an external meter reading source as an example. The process of registering a user or a collection agent is the same.

First, the external meter-reader registers the identifier of the terminal 100 in the external terminal management unit 311 . Here, the identifier may be a mobile phone number or a separately defined format.

Now, the external terminal manager 311 assigns a reference address to the identifier of the registered terminal 100 . Here, the reference address serves as a criterion for classifying an area in which an external meter reading agent performs a meter reading task.

The external terminal management unit 311 allocates an identifier of the terminal 100 to a household number having an address within a predetermined distance based on the reference address. That is, generations are assigned. Although the distance was used as a criterion for allocating households, various criteria such as the total number of households and apartment complexes can be used.

Now, when the terminal 100 requests the meter reading data, only the meter reading data for the household to which the meter reading task is assigned to the user of the terminal 100 using the identifier of the terminal 100 will be confirmed. For example, in the example shown in FIG. 5 , when the terminal 100 of “Hong Gil-dong” requests meter reading data, the household member ?@mak? Only data in which "Hong Gil-dong" is listed is confirmed.

Specifically, when the terminal 100 requests the meter reading data of the meter reading target, the external terminal management unit 311 checks the identifier of the requested terminal 100 , and the meter reading generation management unit 312 allocates the identifier of the terminal 100 . It can be confirmed by transmitting the meter reading data for the household number to the terminal 100, and the terminal 100 outputs the transmitted meter reading data.

3.1 Example 1 - Meter reading work by an external meter reading source

With reference to FIG. 8, the meter reading operation by the external meter reading source of Example 1 is demonstrated.

First, the terminal 100 makes a connection request to the control server 300 . It may be a method such as login.

The external terminal manager 311 of the control server 300 examines whether the identifier of the access-requested terminal 100 is registered, and checks whether the access-requested terminal is a registered terminal (S101).

If it is a registered terminal 100 , the meter reading data is transmitted to the terminal 100 . Only the meter reading data corresponding to the meter reading target is transmitted using the identifier of the terminal 100 (S102).

Now, the external meter reader moves to the corresponding household and approaches the meter (S111) to take an image. Accordingly, a meter reading meter image is generated in the terminal 100 (S121).

The meter reading meter image photographed by the terminal 100 is transmitted to the image analysis AI server 200 (S122).

The image analysis AI server 200 uses the transmitted meter reading meter image to check the object number and the guide value, which is transmitted to the control server 300 (S123).

The household number confirmation unit 313 confirms the household number using the confirmed article number (S124).

The meter-reading household management unit 312 double-checks which household the meter is by using the checked household number and the instrument number in the meter-reading data, and if confirmed, updates the guideline value. At the same time, the meter reading meter image is updated, and the meter reading date and time is updated by checking the time information at which the meter reading meter image is generated ( S125 ). If updated, the meter reading data has the same format as the first generation of FIG. 5 .

The rate calculating unit 320 calculates the gas rate in a preset method using the meter reading data and the guide value (S126).

Meanwhile, in this process, in order to further confirm whether the generation corresponding to the photographed meter is accurate, it may be checked threefold, and for this, GPS information of the terminal 100 may be utilized. The meter reading meter image further includes GPS information of the terminal 100 at the time of generating the image, and since the meter reading data includes an address previously mapped to the household number, the household number check unit 313 checks the identity of the household number to verify the accuracy of In this way, when the meter reading data is stored indefinitely, or when the image analysis AI server 200 erroneously performs an analysis, it is possible to prevent a case in which a charge is erroneously charged.

When the meter reading of one generation meter is completed through this process (S121 to S128), the external meter reading source moves toward the next meter. At this time, the moving path is the meter reading path.

Conventionally, according to the gas billing date that is different for each household, the order of meter reading has been set manually, but the present invention can automate this and guide the next most effective meter reading meter to an external meter reader according to priority.

To this end, the artificial intelligence model built by machine learning in the meter reading sequence suggestion unit 314 is utilized. There are three factors used here: physical distance, time taken during past travel (ie, time taken until the next meter reading), and the route actually selected by an external meter.

The physical distance is a factor for shortening the movement path of the external meter reading source.

The time taken for moving in the past is a factor to consider because the actual travel time may vary depending on the difference in altitude and whether there is a road between moving households even when the physical distance is the same.

The route actually selected by the external meter is a factor to consider other factors, such as the preferred route of the external meter that cannot be considered only by the physical distance and time taken in the past.

Specifically, the meter-reading sequence suggestion unit 314 determines the distance between the address of one household number and the address of another household number among the household numbers included in the meter-reading data, the meter reading date and time input for any one household number, and An artificial intelligence model is generated by performing machine learning by using as input data the difference in meter reading date and time input for different household numbers and the order in which the meter reading was performed for each household number as input data.

If any one household number (i.e., the most recent meter reading and the current location of the external meter reading source) is identified in the model, the shorter the distance, the shorter the meter reading time difference, and the closer the meter reading rank, the higher the score. The score is calculated in the order of meter reading set to The higher the score of the meter reading sequence, the higher the recommendation for the next destination.

Using this model, first, the meter reading sequence calculating unit 314 sets the current generation number indicating the position of the external meter reading source as a reference (S127), and the meter reading meter image is not updated (that is, the meter reading has not been performed yet). The meter reading order score is calculated for all other household numbers, and the household number with the highest meter reading order score is checked and ranked No. 1 (S128) (refer to FIG. 5).

Now, if the previous step is repeated (S131) with the household number assigned the first rank as a new reference (S130), the next second-ranked household number is checked, and this process is repeated to other non-ranked households If it is repeated until there is no household (S129), the ranking of all household numbers requiring meter reading is confirmed. These rankings are the meter reading orders recommended for external meter reading personnel, and if they are followed in numerical order, the meter reading order becomes the order. Of course, the meter reading sequence calculating unit 314 may output the meter reading sequence from the terminal 100 on a map (S132).

According to the prior art, a professional meter reading officer who has been trained to read the guide value must do the meter reading, and it takes a lot of time to read the meter depending on how to directly determine the meter reading sequence. Therefore, anyone can read the meter, and the order of the meter reading is automatically optimized and suggested, so it is more effective and quicker.

3.2 Example 2 - Meter replacement work

With reference to FIG. 9, the meter replacement task of Example 2 is demonstrated.

Unlike Embodiments 1, 3, and 4, the meter replacement task is directly performed by a gas supplier employee rather than an external employee, and a separate terminal confirmation step is not required.

First, replacement meter data is transmitted from the control server 300 to the terminal 100 (S201). In the replacement meter data, a household number, a demolition piece number, an installed piece number, and an address have already been entered (see FIG. 6 ).

The employee has a newly installed meter, checks the address on the replacement meter data, moves to the household, and approaches the demolition meter to be demolished (S211).

When the staff photograph the demolition meter, the demolition meter image is generated in the terminal 100 (S221).

The terminal 100 transmits the demolition meter image to the image analysis AI server 200 (S222), and the image analysis AI server 200 checks the article number and the guide value here and transmits it to the control server 300 (S223). ).

The household number confirmation unit 313 checks the household number by using the confirmed article number (S224).

In the meter replacement management unit 315, whether the item number of the demolition target meter input in advance and the piece number confirmed in the image are the same, and the household number and the household number confirmation unit 313 of the pre-entered meter replacement target household By checking whether the identified household numbers are the same, double check whether the correct meter has been accessed. If all are the same, a demolition meter confirmation signal is transmitted to the terminal 100 (S225). That is, a confirmation signal indicating that it should be demolished has been transmitted.

In this case, the meter replacement management unit 315 specifies the household using the household number and the equipment number in the replacement meter data, and then updates the checked guideline value to the demolition guideline value. The updated demolition guideline value will be used for the next month's fee settlement. In addition, the demolition meter image is updated, and the demolition date and time is updated by checking the time information at which the demolition meter image is generated (S226).

On the other hand, the employee has received the demolition meter confirmation signal through the terminal 100, along with safety measures, the demolition meter is demolished and the installation meter is installed (S231).

When the installation of the installation meter is completed, an image of the installation meter is generated in the terminal 100 by photographing it using the terminal 100 (S241).

Similarly, the terminal 100 transmits the installed meter image to the image analysis AI server 200 (S242), and the image analysis AI server 200 uses the installed meter image to check the article number and guide value, thereby controlling the server It is transmitted to (300) (S243).

The meter replacement management unit 315 checks whether the object number of the meter to be installed in advance input in the replacement meter data is the same as the object number confirmed in the image, and if the same, transmits a replacement confirmation signal to the terminal 100 ( S244). That is, it is confirmed that the replacement is normally performed with an accurate meter.

Now, the meter replacement management unit 315 updates the guide value to the installation guide value by using the household number and the product number in the replacement meter data. In addition, the installation meter image is updated, and information on the time when the installation meter image is generated is checked to update the installation date and time ( S245 ).

These updates may be simultaneously updated to the meter reading data in addition to the replacement meter data. In this case, the next month's bill calculation becomes more convenient. That is, the meter replacement management unit 315 may specify a household by using the household number and the appliance number in the meter reading data, and update the guide value, the replacement meter image, and the meter reading date and time.

Meanwhile, in the present embodiment, it is possible to triple check whether the user visited the correct household by using the GPS information of the terminal 100 .

Specifically, the household number verification unit 313 verifies the accuracy of the identified household number by comparing the GPS information included in the demolished meter image and the address transmitted from the replacement meter data.

According to the prior art, when replacing the meter, an employee visits the household and stores the demolished meter in the ENG (warehouse), separately checking the guide value and transmitting it to the customer center to calculate the rate, so it is difficult to calculate the rate immediately. Although the cumbersome work was done, according to the present invention, the staff can omit all procedures by photographing only the image of the meter before demolition.

3.3 Example 3 - Meter management tasks when moving out

With reference to FIG. 10, the meter management task at the time of moving out of Example 3 is demonstrated.

In this embodiment, users of each household directly perform meter-related tasks. Accordingly, after the user registers his/her terminal 100 with the control server 300 , the user needs to request access to the control server 300 . It may be a method such as member registration and login.

The external terminal management unit 311 of the control server 300 examines whether the identifier of the access-requested terminal 100 is registered, and checks whether the access-requested terminal is a registered terminal (S301).

At the same time, the terminal 100 must transmit the customer identifier and the moving-out processing signal to the moving-out household management unit 316 (S302). This is to inform the control server 300 that the customer applies for moving out.

Now, the user accesses his household meter and takes an image of the meter. Accordingly, an image of a billing meter is generated in the terminal 100 (S321).

The terminal 100 transmits the moving-out meter image to the image analysis AI server 200 (S322), and the image analysis AI server 200 checks the object number and the guide value and transmits it to the control server 300 (S323) .

The household number confirmation unit 313 confirms the household number using the confirmed article number (S324).

The moving-out household management unit 316 checks whether the identified household number is the same as the household number corresponding to the customer identifier transmitted in step S302. In the same case, after specifying the household using the household number and the equipment number in the meter reading data, the guide value is updated, and at the same time, the moving out meter image and the meter reading date and time, which is the time information at which the moving out meter image is generated, are updated (S325).

Now, in the control server 300 of the gas supplier, the fee calculation unit 320 calculates a fee using the checked guide value and transmits it to the terminal 100 (S326). When the user pays the transmitted fee, the work related to moving out is terminated. Practically, the user is done with all the work just by taking an image of the meter of his or her household.

Next, the gas supply control unit 330 generates a signal for stopping the supply of gas transmitted to the identified household number, which is transmitted to the gas supply company server 400 , thereby stopping the supply of gas to the corresponding household. After that, the manager, who is an employee, can visit the household at a convenient time and take safety measures (S331).

Meanwhile, in the present embodiment, it is possible to triple check whether the user visited the correct household by using the GPS information of the terminal 100 .

Specifically, the household number check unit 313 verifies the accuracy of the checked household number by comparing the GPS information included in the moving-out meter image with the address stored in the meter reading data.

According to the prior art, the moving-out person visits the customer center a few days before moving out or applies for moving-out processing through the website or landline phone, and the available staff directly visits the household to check the guide value, and then the customer center again sends the corresponding fee. was decided and notified to the moving-out, and then the moving-out had to go through a complicated and cumbersome moving-out process to pay the fee, but according to the present invention, the moving-in person simply takes a picture of his/her meter and uploads it. .

3.4 Example 4 - Collection work

With reference to FIG. 11, the collection operation of Example 4 is demonstrated.

The collection center requests the terminal 100 access to the control server 300 .

The external terminal manager 311 of the control server 300 checks whether the terminal 100 for which the connection request is made is a registered terminal (S401).

If it is confirmed, the collection data is transmitted to the terminal 100 (S402). Here, it includes meter reading data of a predetermined generation as a collection target.

The collection officer visits the household by verifying the address in the collection data. After the visit, the collection officer may decide that collection is possible or that collection is impossible. If it is determined that collection is impossible, the gas supply is stopped (S411).

To this end, the collection officer shoots the collection meter. That is, a collection meter image is generated in the terminal 100 (S421).

The terminal 100 transmits the collection meter image to the image analysis AI server 200 (S422), and the image analysis AI server 200 uses this to check the article number and the guide value and transmits it to the control server 300 (S423).

The household number confirmation unit 313 confirms the household number using the confirmed article number.

The collection management unit 317 checks whether the identified household number is the same as the household number of the collection data, and if it is the same, updates the guide value for the household specified by using the household number and the article number in the collection data, Update the collection meter image, check the information on the time the collection meter image is generated to update the supply-oriented date and time (S425).

The updated guide value is used to calculate the fee until the gas supply is stopped, and the fee calculating unit 320 calculates the fee using the guide value (S426) and notifies the corresponding household.

At the same time, the gas supply control unit 330 generates a signal for stopping the supply of gas transmitted to the identified household number (S427), which is transmitted to the gas supplier server 400, thereby stopping the supply of gas to the corresponding household. becomes (S431).

Meanwhile, in the present embodiment, it is possible to triple check whether the user visited the correct household by using the GPS information of the terminal 100 .

Specifically, the household number verification unit 313 verifies the accuracy of the verified household number by comparing the GPS information included in the collection meter image with the address stored in the meter reading data.

According to the prior art, when the collection agency conducts collection activities, even if the collection officer visits the household and determines that collection is impossible, the customer center is notified of this separately, and the customer center directly visits the household and checks the guideline value. , again, the customer center had to go through a complicated and cumbersome process to determine the corresponding fee. can be completed

100: terminal
200: image analysis AI server
290: object image database
300: control server
311: external terminal management unit
312: meter reading household management unit
313: household number verification unit
314: meter reading order calculation unit
315: meter replacement management unit
316: Moving out household management department
320: fee management department
330: gas supply control unit
391: Generation Database
392: Meter reading sequence database
393: Meter database
394: collection database
400: gas supplier server

Claims (13)

(a1) checking, by the control server 300, whether the terminal 100 requested to access the control server 300 is a registered terminal;
(b1) transmitting, by the control server 300, the meter reading data of the meter reading target to the terminal 100; - The meter reading data includes a household number and an object number mapped thereto -
(c1) generating a meter reading meter image in the terminal 100;
(d1) transmitting, by the terminal 100, an image of the meter reading meter to the image analysis AI server 200;
(e1) transmitting, by the image analysis AI server 200, to the control server 300 by using the meter reading meter image transmitted in the step (d1) to check the article number and the guide value;
(f1) confirming the household number by the household number confirmation unit 313 of the control server 300 using the unit number confirmed in the step (e1);
(g1) the meter-reading household management unit 312 updates the guideline value using the household number and the article number confirmed in step (f1) in the meter-reading data, updates the meter-reading meter image, and the meter-reading meter image checking the generated time information to update the meter reading date and time; and
(h1) calculating, by the rate calculating unit 320, the gas rate using the meter reading data and the guide value updated in the step (g1),
The meter reading sequence suggestion unit 314 is configured to provide a distance between an address of one household number and an address of another household number among the household numbers included in the meter reading data, and a household different from the meter reading date and time input for any one household number. By performing machine learning using the difference in the date and time of the meter reading entered for the number and the order in which the meter reading was made by each household number as input data, machine learning is performed. The score is calculated in the order of meter reading set so that the shorter and closer the meter reading rank is, the larger the score.
After step (g1),
(i11) setting, by the meter reading sequence calculating unit 314, based on the generation number confirmed in step (f1);
(i12) The meter-reading sequence calculating unit 314 calculates the meter-reading sequence score for all other household numbers for which the meter-reading meter image is not updated, using the criteria, and identifies the household number with the highest meter-reading sequence score and ranks it granting;
(i13) The meter reading sequence calculation unit 314 continuously repeats the step (i12) based on the ranked household number as a new standard, so that all other household numbers for which the meter reading meter image is not updated are ranked. and calculating a meter reading order using the assigned rank; and
(i14) the step of outputting, by the meter reading sequence calculating unit 314, the meter reading sequence on a map,
Way.
The method of claim 1,
The meter reading meter image further includes GPS information of the terminal 100 at the time of image creation, and the meter reading data further includes an address previously mapped to a household number,
The step (f1) is,
The household number check unit 313 compares the GPS information included in the meter reading meter image with the address stored in advance in the meter reading data as corresponding to the household number confirmed in step (f1), thereby confirming the household verifying the correctness of the number;
Way.
delete The method of claim 1,
Before step (a1),
(a01) the external terminal management unit 311 receiving the identifier of the terminal 100 from the terminal 100 and registering it;
(a02) assigning, by the external terminal management unit 311, a reference address to the identifier of the registered terminal 100; and
(a03) the external terminal management unit 311 allocating an identifier of the terminal 100 to a household number having an address within a predetermined distance based on the reference address,
The meter reading data of the meter reading target transmitted to the terminal 100 in the step (b1) is data on the household number assigned in the step (a03) with respect to the identifier of the terminal 100,
Way.
5. The method of claim 4,
(a04) requesting, by the terminal 100, the meter reading data of the meter reading target;
(a05) confirming, by the external terminal management unit 311, the identifier of the terminal 100 requested in the step (a04);
(a06) transmitting, by the meter reading generation management unit 312, meter reading data for the generation number to which the identifier of the terminal 100 is assigned to the terminal 100; and
(a07) outputting the transmitted meter reading data by the terminal 100,
The generation information output in step (a07) is,
Includes household number, equipment number, meter reading date and time, guide value, address, meter reading person, meter reading rank and image link,
Here, the meter reading date and time is the most recent input as the meter reading date and time input in the step (g1), the meter reading source is assigned to each identifier of the terminal 100 confirmed in the step (a05), and the meter reading order is the (i12) This is the ranking confirmed in step, and when the address included in the household information is activated, the address is output together with the map, and when the image link included in the household information is activated, the meter reading image transmitted in step (d1) is output felled,
Way.
The method of claim 1,
After step (h1),
(b21) transmitting the household number of the household to be replaced and the address assigned thereto from the control server 300 to the terminal 100;
(c21) generating a demolition meter image in the terminal 100;
(d21) transmitting, by the terminal 100, a demolition meter image to the image analysis AI server 200;
(e21) the image analysis AI server 200, using the demolition meter image transmitted in the step (d21) to check the object number and the guide value and transmit it to the control server 300;
(f21) confirming the household number by the household number confirmation unit 313 of the control server 300 using the unit number confirmed in the step (e21);
(g21) the meter replacement management unit 315 determines whether the pre-input number of the demolition target meter is the same as the piece number confirmed in step (e21), and the pre-entered household number of the meter replacement target household Checking whether the household numbers confirmed in step (f21) are the same, and transmitting a demolition meter confirmation signal to the terminal 100 if they are all the same; and
(h21) The meter replacement management unit 315 uses the household number confirmed in the confirmed step (f21) and the unit number confirmed in the step (e21) in the replacement meter data, and sets the guide value to the removal guideline updating the value, updating the demolition meter image, and updating the demolition date and time by confirming the time information at which the demolition meter image was generated,
Way.
7. The method of claim 6,
After the step (h21),
(c22) generating an installed meter image in the terminal 100;
(d22) transmitting, by the terminal 100, an installed meter image to the image analysis AI server 200;
(e22) the image analysis AI server 200, using the installed meter image transmitted in the step (d22) to check the equipment number and the guide value and transmit it to the control server 300;
(g22) The meter replacement management unit 315 checks whether the equipment number of the meter to be installed in advance and the equipment number confirmed in step (e22) are the same, and if they are the same, the replacement confirmation is sent to the terminal 100 transmitting a signal; and
(h22) the meter replacement management unit 315 uses, in the replacement meter data, the household number confirmed in the confirmed step (f22) and the unit number confirmed in the step (e22), in the step (e22) The method further comprising the steps of updating the determined guideline value to the installation guideline value, updating the installation meter image, and updating the installation date and time by checking information about the time when the installation meter image was created,
Way.
8. The method of claim 7,
The demolition meter image further includes GPS information of the terminal 100 at the time of image creation,
The step (f21) is,
Comprising the step of verifying, by the household number verification unit 313, the accuracy of the identified household number by comparing the GPS information included in the demolished meter image with the address transmitted in the step (b21),
Way.
9. The method of claim 8,
The step (h22) is,
The meter replacement management unit 315 updates the guideline value by using the household number confirmed in the step (f21) and the appliance number confirmed in the step (e22) in the meter reading data, and updates the installed meter image, , further comprising the step of updating the meter replacement date and time confirmed in step (h22) to the meter reading date and time,
Way.
The method of claim 1,
After step (h1),
(b3) transmitting a customer identifier and a moving-out processing signal from the terminal 100 to the moving-out household management unit 316;
(c3) generating a billing meter image in the terminal 100;
(d3) transmitting, by the terminal 100, an image of a transfer meter to the image analysis AI server 200;
(e3) the image analysis AI server 200, using the transfer meter image transmitted in the step (d3) to check the equipment number and guide value and transmit it to the control server 300;
(f3) confirming the household number by the household number verification unit 313 of the control server 300 using the unit number confirmed in the step (e3);
(g3) the moving-out household management unit 316 checks whether the household number confirmed in the step (f3) is the same as the household number corresponding to the customer identifier transmitted in the step (b3), and if the same, the meter reading data , updating a guideline value by using the identified household number and appliance number, updating the moving-out meter image, and checking the time information at which the moving-out meter image was generated to update the meter reading date and time;
(h3) the fee calculating unit 320 calculating a fee using the updated guide value in step (g3) and transmitting the calculated fee to the terminal 100; ??
(i3) comprising the step of generating, by the gas supply control unit 330, a signal for stopping the gas supply transmitted to the identified household number,
Way.
11. The method of claim 10,
The moving-out meter image further includes GPS information of the terminal 100 at the time of image creation, and the meter reading data further includes an address previously mapped to a household number,
The step (f3) is,
The household number confirmation unit 313 verifies the accuracy of the confirmed household number by comparing the GPS information included in the moving-out meter image with the address previously stored in the meter reading data as corresponding to the identified household number. comprising the steps of
Way.
The method of claim 1,
After step (h1),
(b4) transmitting, by the control server 300, collection data including meter reading data of a predetermined generation as a collection target to the terminal 100; - The collection data includes a household number and an object number mapped thereto -
(c4) generating a collection meter image in the terminal 100;
(d4) transmitting, by the terminal 100, a collection meter image to the image analysis AI server 200;
(e4) the image analysis AI server 200, using the collection meter image transmitted in the step (d4) to check the article number and the guide value and transmitting to the control server 300;
(f4) confirming, by the household number confirmation unit 313 of the control server 300, the household number using the unit number confirmed in the step (e4);
(g4) the collection management unit 317 checks whether the household number confirmed in step (f4) is the same as the household number transmitted in step (b4), and if the same, in the collection data, the identified household Updating the guideline value using the number and the article number, updating the collection meter image, and checking the time information at which the collection meter image was generated to update the supply-oriented date and time;
(h4) calculating, by the rate calculating unit 320, a rate using the updated guide value in step (g4); and
(i4) comprising the step of generating, by the gas supply control unit 330, a signal for stopping the supply of gas transmitted to the identified household number,
Way.
13. The method of claim 12,
The collection meter image further includes GPS information of the terminal 100 at the time of image creation, and the collection data further includes an address mapped in advance to a household number,
The step (f4) is,
The household number confirmation unit 313 verifies the accuracy of the identified household number by comparing the GPS information included in the collection meter image and the address stored in advance in the collection data as corresponding to the identified household number. comprising the steps of
Way.

Images