Artificial intelligence-based collection server and its operation method for collecting meter readings in conjunction with camera arranged to capture meter reading display area of meter

Abstract

To provide an artificial intelligence-based collection server and its operation method that enable the collection of meter readings of a meter more conveniently and efficiently, without requiring a meter reader to directly obtain meter readings from the meter.SOLUTION: A collection server 110 includes: a collection event generation unit that generates a collection event for collecting meter readings at preset collection cycle intervals; a capture request unit that transmits a capture request command via the collection event generation unit, instructing a camera to capture and transfer an image of the area on the meter where the meter reading is displayed; an identification unit that, upon the transfer of the captured image of the area on the meter where the meter reading is displayed from the camera in response to the capture request command, identifies a first meter reading composed of numerical values from the captured image based on a preset number identification model; and a history storage processing unit that, after generating meter reading collection history information at the time the collection event occurred, stores the meter reading collection history information in a history database in which multiple pieces of meter reading collection history information are stored.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an artificial intelligence-based collection server that collects measurement values ​​in conjunction with a camera arranged to photograph an area where measurement values ​​are displayed on a weighing scale, and an operating method thereof. [Background technology]

[0002] Typically, in homes and industrial sites where demand for electricity, gas, water, etc. is generated, various types of meters are installed to measure the amount of electricity, gas, water, etc. used.

[0003] Recently, digital weighing instruments have appeared that can transmit measured values ​​to a remote system management device via a predetermined communication module, providing convenience to users.

[0004] As such, a variety of digital meters equipped with designated communication modules that can transmit measurement values ​​from remote locations have appeared, but until now, there have been many meters that require meter readers to personally visit the area where the meter is installed and read the measurement values ​​with the naked eye.

[0005] As a result, labor costs are incurred as meter readers personally go around and read meters, and there is also distrust in the differences in charges due to errors that occur when meter readers read meters.

[0006] Therefore, there is a need to introduce a system that will enable meter readers to collect the measurement values ​​of meters that do not have a communication module from remote locations without the need for direct meter reading.

[0007] In this regard, if a system were introduced that works in conjunction with a camera positioned to photograph the area on the meter where the measurement value is displayed, receives an image of the measurement value from the camera, and can identify the measurement value from the received image, it would be possible to collect the measurement value of the meter more conveniently and efficiently, without the need for meter readers to collect the measurement value directly.

[0008] Therefore, there is a need for research into technology related to a system for collecting metric values ​​in conjunction with a camera positioned to photograph the area of ​​the scale where the metric values ​​are displayed. Summary of the Invention [Problem to be solved by the invention]

[0009] The present invention provides an artificial intelligence-based collection server and its operating method that collects measurement values ​​in conjunction with a camera positioned to photograph the area where the measurement values ​​are displayed on a meter, thereby eliminating the need for meter readers to collect the measurement values ​​themselves and making it possible to collect meter measurement values ​​more conveniently and efficiently. [Means for solving the problem]

[0010] According to an embodiment of the present invention, an AI-based collection server that collects measurement values ​​in conjunction with a camera disposed to capture an image of an area on a meter where the measurement values ​​are displayed includes a collection event generating unit that generates a collection event for collecting measurement values ​​at a predetermined collection cycle interval; an image requesting unit that, when the collection event occurs in any cycle via the collection event generating unit, transmits an image requesting command to request the camera to capture and transfer an image of the area on the meter where the measurement values ​​are displayed; and a collection event generating unit that receives an image requesting command from the camera and transmits the image of the area on the meter where the measurement values ​​are displayed. When a captured image of a user is transferred, the system includes an identification unit that identifies a first metric value consisting of numbers from the captured image based on a preset digit identification model (the digit identification model is an artificial intelligence-based model that has been machine-learned in advance to identify digits present in the image and calculate an identification result based on the digits) and a history storage processing unit that generates date / time information when the collection event occurred, generates metric value collection history information consisting of the date / time information and the first metric value, and then stores the metric value collection history information in a history database in which multiple metric value collection history information is stored.

[0011] In addition, according to one embodiment of the present invention, a method for operating an AI-based collection server that collects measurement values ​​in conjunction with a camera disposed to capture an image of an area on a weighing scale where the measurement values ​​are displayed includes the steps of: generating a collection event for collecting measurement values ​​at a predetermined collection cycle interval; when the collection event occurs in any cycle through the step of generating the collection event, transmitting a photography request command requesting the camera to capture and transfer an image of the area on the weighing scale where the measurement values ​​are displayed; and transmitting the image of the area on the weighing scale where the measurement values ​​are displayed from the camera in response to the photography request command. When a captured image of the indicated area is transferred, the method includes the steps of identifying a first metric value consisting of numbers from the captured image based on a pre-set digit identification model (the digit identification model means an artificial intelligence-based model that has been pre-trained through machine learning to identify digits present in the image and calculate an identification result based on the digit identification model); and generating date / time information of the time when the collection event occurred, generating metric value collection history information consisting of this date / time information and the first metric value, and then storing the metric value collection history information in a history database in which multiple metric value collection history information is stored. [Effects of the Invention]

[0012] The present invention provides an artificial intelligence-based collection server and its operating method that collects measurement values ​​in conjunction with a camera positioned to photograph the area where the measurement values ​​on a meter are displayed, thereby eliminating the need for meter readers to collect the measurement values ​​of the meter directly, making it possible to collect the measurement values ​​of the meter more conveniently and efficiently. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 2 is a diagram illustrating a configuration of a collection server according to an embodiment of the present invention. [Figure 2] 4 is a diagram illustrating the operation of a collection server according to an embodiment of the present invention. FIG. [Figure 3] 4 is a diagram illustrating the operation of a collection server according to an embodiment of the present invention. FIG. [Figure 4] 3 is a flowchart illustrating a method of operation of a collection server according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Such description is not intended to limit the present invention to a specific embodiment, but includes any modifications, equivalents, or alternatives within the spirit and technical scope of the present invention. While describing each drawing, like reference numerals are used for like components, and unless otherwise specified, all terms used in this specification, including technical or scientific terms, have the same meaning as commonly understood by those having ordinary skill in the art to which the present invention belongs.

[0015] In this specification, when a part "comprises" a certain element, this does not mean that other elements are excluded, but that other elements may also be included, unless otherwise specified. Furthermore, in various embodiments of the present invention, each element, functional block, or means may be configured as one or more sub-elements, and the electrical, electronic, or mechanical functions performed by each element may be embodied by any known element or mechanical element, such as an electronic circuit, an integrated circuit, or an ASIC (Application Specific Integrated Circuit), and may be embodied individually or by integrating two or more elements into one.

[0016] The blocks in the accompanying block diagrams and steps in the flowcharts may be interpreted as computer program instructions that are loaded into a processor or memory of a data processing device, such as a general-purpose computer, a special-purpose computer, a portable personal computer, or a network computer, and that execute the specified functions. These computer program instructions may be stored in a memory or computer-readable memory of a computer device, so that the functions described in the blocks in the block diagrams or steps in the flowcharts may also be implemented as a product containing instruction means for executing the functions. Each block or step may represent a module, segment, or part of code that includes one or more executable instructions for executing specific logical function(s). It should also be noted that in some alternative embodiments, the functions described in the blocks or steps may be executed out of the specified order. For example, two blocks or steps shown in succession may be executed substantially simultaneously, may be executed in reverse order, or may even omit some blocks or steps.

[0017] FIG. 1 is a diagram showing the configuration of a collection server according to an embodiment of the present invention.

[0018] Referring to FIG. 1, the collection server 110 according to the present invention is an artificial intelligence-based collection server that collects measurement values ​​in conjunction with a camera 20 arranged to photograph the area where measurement values ​​are displayed on a meter 10, and is composed of a collection event generation unit 111, a photography request unit 112, an identification unit 113, and a history storage processing unit 114.

[0019] The collection event generating unit 111 generates a collection event for collecting measurement values ​​at a preset collection cycle interval. For example, if the collection cycle is set to "one hour," the collection event generating unit 111 can generate a collection event for collecting measurement values ​​every "one hour."

[0020] When the collection event occurs in any cycle through the collection event generating unit 111, the photography requesting unit 112 transmits a photography request command requesting the camera 20 to photograph and transfer the area where the measurement value is displayed on the weighing scale 10.

[0021] Then, in response to the image capture request command, the camera 20 captures an image of the area on the weighing scale 10 where the measurement value is displayed, generates a captured image as shown by reference numeral 211 in FIG. 2, and then transmits the captured image to the collection server 110 of the present invention.

[0022] In this way, when a captured image of the area where the measurement value is displayed on the weighing scale 10 is transmitted from the camera 20 to the collection server 110 of the present invention in response to the image capture request command, the identification unit 113 identifies a first measurement value consisting of a number from the captured image based on a preset number identification model. Here, the number identification model refers to an artificial intelligence-based model that has been trained by machine learning in advance to identify numbers present in an image from the image and calculate an identification result accordingly. In this case, the number identification model may be embodied as a classification model based on a convolutional neural network (CNN) that passes an input image through a convolutional layer to construct a predetermined feature map, passes the feature map through a fully-connected layer, and classifies the number corresponding to the feature map from among multiple numbers.

[0023] In this case, according to an embodiment of the present invention, when the captured image is transmitted from the camera 20, the identification unit 113 detects a rectangular box image of an object corresponding to an area where a metric value exists from the captured image based on a preset object detection model, and then inputs and applies the box image to the digit identification model to identify the first metric value consisting of a digit from the box image. Here, the object detection model refers to a machine-learned artificial intelligence-based model for detecting an object corresponding to an area where a metric value exists from an image in the form of a rectangular box. In this regard, the object detection model may be embodied as a YOLO (You Only Look Once)-based object detection model for detecting an object corresponding to an area where a metric value exists.

[0024] The process in which the identification unit 113 identifies a metric value from the captured image will be described below with reference to FIG.

[0025] First, it is assumed that the captured image received from the camera 20 is the same as the image indicated by reference numeral 211. Then, the identification unit 113 applies the captured image as an input to a YOLO-based object detection model as indicated by reference numeral 212, thereby detecting a rectangular box image 210 for an object corresponding to an area where a metric value exists as indicated by reference numeral 213.

[0026] Then, the identification unit 113 applies the box image 210 as an input to the number identification model as shown in reference numeral 214, thereby being able to identify the first metric value consisting of numbers, "0072241", from the box image 210 as shown in reference numeral 215.

[0027] In this case, according to one embodiment of the present invention, the identification unit 113 may further include a configuration for configuring the box image as a binary image and applying it as an input to the digit identification model so that digit identification from the box image can be performed more accurately.

[0028] In this regard, when the box image is detected, the identification unit 113 converts pixels having pixel values ​​according to a hue that belongs to a first hue range that is predetermined for the color of the number among the plurality of pixels constituting the box image into a predetermined first pixel value, and converts the pixel values ​​of the remaining pixels into a predetermined second pixel value, thereby converting the box image into a binary image consisting of pixels having the first pixel value and the second pixel value, and then applies the box image converted into a binary image to the number identification model as an input, thereby being able to identify the first metric value consisting of a number from the box image.

[0029] In this regard, as in the above example, if the box image is the same as the one shown in the drawing reference numeral 210, and the range of the first hue is "R value: 200 to 255, G value: 200 to 255, B value: 200 to 255", which are values ​​corresponding to the white series when based on RGB values, and the range of the second hue is "R value: 0 to 50, G value: 0 to 50, B value: 0 to 50", which are values ​​corresponding to the black series when based on RGB values, and the first pixel value is "R value: 255, G value: 255, B value: 255", which are values ​​corresponding to the black series when based on RGB values. " and when the second pixel value is based on RGB values, it is "R value: 0, G value: 0, B value: 0", the identification unit 113 converts pixels having pixel values ​​according to a hue belonging to the first hue range among the plurality of pixels constituting the box image 210 to the first pixel value, and converts pixel values ​​of the remaining pixels to the second pixel value, thereby converting the box image 210 into a binary image 310 consisting of pixels having the first pixel value and the second pixel value, as shown by reference numeral 310 in FIG. 3.

[0030] In this way, when the box image 210 is converted into the binary image 310, the numbers and other parts present on the box image 210 are clearly distinguished in color from the first pixel value and the second pixel value, so that the accuracy of number identification performed from the box image 210 can be significantly improved.

[0031] When the identification of the first metric value from the captured image acquired from camera 20 is completed via identification unit 113 according to the method described above, history storage processing unit 114 generates date / time information of the time when the collection event occurred, generates metric value collection history information consisting of this date / time information and the first metric value, and then stores the metric value collection history information in history database 115, which stores multiple metric value collection history information.

[0032] For example, if the collection event generation unit 111 determines that the time when the collection event occurred was "13:25:30, March 26, 2024" and the first measurement value was identified as "0072241," the history storage processing unit 114 can generate measurement value collection history information in the form shown in Table 1 below.

[0033] [Table 1]

[0034] Then, the history storage processing unit 114 can further store the metric value collection history information such as that in Table 1 above in the history database 115 in which multiple previously generated metric value collection history information is stored.

[0035] According to an embodiment of the present invention, the collection server 110 may further include a history information provider 116 .

[0036] When the history information providing unit 116 receives a command from a pre-designated administrator terminal 30 requesting the provision of metric value collection history information for a first period designated by the administrator, it extracts first metric value collection history information having date / time information belonging to the first period from the multiple metric value collection history information stored in the history database 115 and transmits it to the administrator terminal 30.

[0037] For example, if the first period specified by the administrator is "January 1, 2024, 10:20:10 to January 20, 2024, 13:30:25," the history information providing unit 116 can extract first metric collection history information having date / time information belonging to the first period, "January 1, 2024, 10:20:10 to January 20, 2024, 13:30:25," from the multiple pieces of metric collection history information stored in the history database 115, and transmit the extracted first metric collection history information to the administrator terminal 30. In this way, the administrator can obtain metric collection history information for the desired period.

[0038] In this case, according to one embodiment of the present invention, the history information providing unit 116 may further include a configuration that, when providing metric value collection history information to the administrator terminal 30, encrypts the history information and provides it to the administrator terminal 30, thereby preventing the metric value collection history information from being exposed to a third party.

[0039] In this regard, the history information providing unit 116 may include a function storage unit 117 , an extraction unit 118 , an encryption event generating unit 119 , a random number generating unit 120 , a hash value generating unit 121 , and an information transmitting unit 122 .

[0040] The function storage unit 117 stores a pseudo-random number generation function that is shared with the administrator terminal 30 in advance.

[0041] When the extraction unit 118 receives from the administrator terminal 30 a command requesting the provision of metric value collection history information for the first period specified by the administrator, it checks the first metric value collection history information having date / time information belonging to the first period from among the multiple metric value collection history information stored in the history database 115, and extracts the first metric value collection history information from the history database 115.

[0042] When the first metric collection history information is extracted, the encryption event generating unit 119 encrypts the first metric collection history information and generates an encryption event for transmitting the first metric collection history information to the administrator terminal 30 .

[0043] When the encryption event occurs, the random number generation unit 120 applies a seed value as input to the pseudo-random number generation function and repeatedly executes the process of generating random numbers n times (n is a natural number greater than or equal to 4).In the first random number generation process, a randomly generated initial seed value is applied as input to the pseudo-random number generation function, and from the second random number generation process onwards, a total of n random numbers are generated by executing the process of specifying the random number generated in the previous step as a seed value and applying it as input to the pseudo-random number generation function.

[0044] For example, if n is 5, the random number generator 120 can generate "random number 1" by applying a randomly generated initial seed value as an input to the pseudo-random number generation function. Then, the random number generator 120 can generate "random number 2" by applying "random number 1" as an input to the pseudo-random number generation function again. Then, the random number generator 120 can generate "random number 3" by applying "random number 2" as an input to the pseudo-random number generation function again. Then, the random number generator 120 can generate "random number 4" by applying "random number 3" as an input to the pseudo-random number generation function again. Finally, the random number generator 120 can generate "random number 5" by applying "random number 4" as an input to the pseudo-random number generation function again. In this manner, the random number generator 120 can generate a total of five random numbers: "random number 1, random number 2, random number 3, random number 4, and random number 5."

[0045] When the hash value generation unit 121 sorts the n random numbers in ascending order and calculates the quartiles, it selects random numbers from the n random numbers that have a magnitude equal to or greater than the first quartile and equal to or less than the third quartile, sequentially concatenates the selected random numbers to generate a concatenated value, and then applies the concatenated value as input to a pre-set hash function (the hash function is also stored in the administrator terminal) to generate a hash value.

[0046] Here, the quartiles are obtained by dividing data sorted in ascending order into four equal parts, with the first quartile being the variable value corresponding to the 1 / 4th place and the third quartile being the variable value corresponding to the 3 / 4th place.

[0047] For example, if the random number generation unit 120 generates seven random numbers, "19, 32, 16, 50, 30, 22, 25," the hash value generation unit 121 rearranges the seven random numbers in ascending order to form "16, 19, 22, 25, 30, 32, 50," and then calculates the quartiles, thereby confirming that the first quartile is "19" and the third quartile is "32."

[0048] Thereafter, the hash value generation unit 121 can select "19, 22, 25, 30, 32" from the seven random numbers as random numbers having a magnitude equal to or greater than the first quartile and equal to or less than the third quartile.

[0049] Then, the hash value generation unit 121 generates a concatenated value of "1922253032" by sequentially concatenating the selected random numbers in order of size, and then applies the concatenated value as an input to a pre-set hash function to generate a hash value.

[0050] In this way, once the generation of the hash value is completed, the information transmission unit 122 encrypts the first metric value collection history information based on the hash value, and then transmits the encrypted first metric value collection history information and the initial seed value to the administrator terminal 30.

[0051] In this case, according to one embodiment of the present invention, when the administrator terminal 30 receives the encrypted first metric collection history information and the initial seed value from the collection server 110 of the present invention, it applies the seed value as an input to the pseudo-random number generation function stored in the administrator terminal 30 and executes the process of generating random numbers repeatedly n times, and in the first random number generation process, it applies the initial seed value as an input to the pseudo-random number generation function, and from the second random number generation process, it executes the process of specifying the random number generated in the previous step as a seed value and applying it as an input to the pseudo-random number generation function, thereby generating the n random numbers.

[0052] Then, when the administrator terminal 30 sorts the n random numbers in ascending order and calculates the quartiles, it selects from the n random numbers those random numbers having a magnitude equal to or greater than the first quartile and equal to or less than the third quartile, sequentially concatenates the selected random numbers to generate a concatenated value, and then applies the concatenated value as input to the hash function stored in the administrator terminal 30 to generate the hash value, and then decrypts the encrypted first metric value collection history information based on the hash value.

[0053] In a related example, if the n random numbers are "19, 32, 16, 50, 30, 22, 25" as in the example described above, when the administrator terminal 30 sorts the n random numbers in ascending order to calculate the quartiles, it selects "19, 22, 25, 30, 32" from the n random numbers as random numbers having a magnitude greater than the first quartile and less than the third quartile, and sequentially concatenates the selected random numbers to generate a concatenated value of "1922253032".Then, it applies the concatenated value as input to the hash function stored in the administrator terminal 30 to generate the hash value, and then decrypts the encrypted first metric value collection history information based on the hash value.

[0054] This allows the administrator to safely obtain the first metric collection history information that belongs to the first period designated by the administrator.

[0055] FIG. 4 is a flowchart illustrating an operation method of an artificial intelligence-based collection server that collects measurement values ​​in conjunction with a camera arranged to capture an area where measurement values ​​are displayed on a weighing scale, according to an embodiment of the present invention.

[0056] In step S410, a collection event is generated to collect measurement values ​​at preset collection cycle intervals.

[0057] In step S420, if the collection event occurs in any cycle through step S410, a photographing request command is transmitted to request the camera to photograph and transmit the area on the meter where the measurement value is displayed.

[0058] In step S430, when a captured image of the area on the weighing scale where the measurement value is displayed is transferred from the camera in response to the image capture request command, a first measurement value consisting of a number is identified from the captured image based on a pre-set number identification model (the number identification model refers to an artificial intelligence-based model that has been machine-learned in advance to identify numbers present on the image and calculate the identification result accordingly).

[0059] In step S440, date / time information at the time the collection event occurred is generated, and then metric value collection history information consisting of this date / time information and the first metric value is generated, and the metric value collection history information is then stored in a history database in which multiple metric value collection history information is stored.

[0060] In this case, according to one embodiment of the present invention, in step S430, when the captured image is transferred from the camera, a rectangular box image for an object corresponding to an area where a metric value exists is detected from the captured image based on a pre-set object detection model (the object detection model means a pre-machine-trained artificial intelligence-based model for detecting an object corresponding to an area where a metric value exists in an image in the form of a rectangular box), and then the box image is applied as an input to the digit identification model, and the first metric value consisting of a digit can be identified from the box image.

[0061] In this case, according to one embodiment of the present invention, in step S430, when the box image is detected, among the plurality of pixels constituting the box image, pixels having pixel values ​​according to a hue belonging to a first hue range pre-specified by the hue of numbers are converted to a pre-set first pixel value, and the pixel values ​​of the remaining pixels are converted to a pre-set second pixel value, thereby converting the box image into a binary image consisting of pixels having the first pixel values ​​and the second pixel values. Then, the box image converted into a binary image is applied as an input to the number identification model, and the first metric value consisting of numbers can be identified from the box image.

[0062] In addition, according to one embodiment of the present invention, the operation method of the collection server may further include a step of extracting first metric value collection history information having date / time information belonging to the first period from among multiple metric value collection history information stored in the history database when a command requesting the provision of metric value collection history information for a first period designated by the administrator is received from a pre-designated administrator terminal, and transmitting the first metric value collection history information to the administrator terminal.

[0063] In this case, according to one embodiment of the present invention, the step of transmitting to the administrator terminal includes a step of holding a function storage unit in which a pseudo-random number generation function, which is shared with the administrator terminal in advance, is stored; a step of, when a command to request provision of metric value collection history information for the first period designated by the administrator is received from the administrator terminal, checking first metric value collection history information having date / time information belonging to the first period among a plurality of metric value collection history information stored in the history database, and extracting the first metric value collection history information from the history database; a step of generating an encryption event for encrypting the first metric value collection history information and transmitting it to the administrator terminal; and a step of, when the encryption event is generated, repeatedly executing a process of applying a seed value as an input to the pseudo-random number generation function and generating a random number n times (n is a natural number equal to or greater than 4), and in the first random number generation process, applying a randomly generated initial seed value to the pseudo-random number generation function. and in the second random number generation step, the random number generated in the previous step is designated as a seed value and applied as an input to the pseudo-random number generation function, thereby generating a total of n random numbers; sorting the n random numbers in ascending order to calculate quartiles, selecting random numbers having a magnitude equal to or greater than the first quartile and equal to or less than the third quartile from the n random numbers, sequentially concatenating the selected random numbers to generate concatenated values, and applying the concatenated values ​​as input to a predetermined hash function (the hash function is also stored in the administrator terminal) to generate a hash value; and encrypting the first metric collection history information based on the hash value, and transmitting the encrypted first metric collection history information and the initial seed value to the administrator terminal.

[0064] In this case, when the administrator terminal receives the encrypted first metric value collection history information and the initial seed value from the collection server, it applies the seed value as an input to the pseudo-random number generation function stored in the administrator terminal and repeatedly executes the process of generating random numbers n times. In the first random number generation process, it applies the initial seed value as an input to the pseudo-random number generation function, and from the second random number generation process, it executes the process of specifying the random number generated in the previous step as a seed value and applying it as an input to the pseudo-random number generation function. Thus, after generating the n random numbers, when the n random numbers are sorted in ascending order to calculate quartiles, it selects random numbers from the n random numbers that have a magnitude that is greater than the first quartile and less than the third quartile, sequentially concatenates the selected random numbers to generate a concatenated value, and then applies the concatenated value as an input to the hash function stored in the administrator terminal to generate the hash value. After that, it can decrypt the encrypted first metric value collection history information based on the hash value.

[0065] The operation method of the collection server according to an embodiment of the present invention has been described above with reference to Fig. 4. Here, the operation method of the collection server according to an embodiment of the present invention can correspond to the configuration for the operation of the collection server 110 described with reference to Figs. 1 to 3, so a detailed description thereof will be omitted.

[0066] The method for operating the collection server according to an embodiment of the present invention may be implemented as a computer program stored in a storage medium for execution through connection with a computer.

[0067] Furthermore, a method for operating a collection server according to an embodiment of the present invention may be embodied in the form of computer program instructions to be executed through connection with a computer and recorded as a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be specially designed and constructed for the present invention, or may be known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include not only machine language code, such as that produced by a compiler, but also high-level language code that can be executed by a computer using an interpreter, etc.

[0068] As mentioned above, the present invention has been described using specific details such as specific components and limited embodiments and drawings, but this is provided merely to assist in a more general understanding of the present invention, and the present invention is not limited to the above-mentioned embodiments. Any modifications and variations can be made from these descriptions by a person having ordinary knowledge in the field to which the present invention pertains.

[0069] Therefore, the spirit of the present invention should not be limited to the described embodiments, and not only the scope of the claims below, but also all equivalent or similar modifications to the scope of the claims belong to the spirit of the present invention. [Explanation of symbols]

[0070] 110 Collection Server 111...Collection event generation section 112 ···Photography Request Department 113 Identification unit 114 History storage processing unit 115 ···History Database 116 ···History Information Department 117 Function storage section 118...Extraction part 119 Encryption Event Generation Section 120 Random number generator 121 Hash value generator 122 Information Transmission Unit 10...Measuring instrument 20 Camera 30 Administrator terminal

Claims

1. An artificial intelligence-based collection server that collects measurement values ​​in conjunction with a camera arranged to capture an area where measurement values ​​are displayed on a measuring instrument, a collection event generating unit that generates a collection event for collecting measurement values ​​at predetermined collection cycle intervals; an image capturing request unit that transmits an image capturing request command to request the camera to capture and transmit an image of an area on the meter where a measurement value is displayed when the collection event occurs in any cycle via the collection event generating unit; an identification unit that, in response to the image capture request command, transfers a captured image of the area of ​​the weighing scale where the measurement value is displayed from the camera, and identifies a first measurement value consisting of a number from the captured image based on a preset number identification model (the number identification model is an artificial intelligence-based model that has been machine-learned in advance to identify numbers present in the image and calculate an identification result therefrom); and and a history storage processing unit that generates date / time information of the time when the collection event occurred, generates metric value collection history information consisting of the date / time information and the first metric value, and then stores the metric value collection history information in a history database in which multiple pieces of metric value collection history information are stored. A collection server comprising:

2. The identification unit When the captured image is transferred from the camera, a rectangular box image of an object corresponding to the area where a metric value exists is detected from the captured image based on a preset object detection model (the object detection model is a model based on artificial intelligence that has been machine-learned in advance to detect an object in the shape of a rectangular box from an image, the object corresponding to the area where a metric value exists), and the box image is then applied as an input to the digit identification model to identify the first metric value consisting of a digit from the box image. The collection server of claim 1 .

3. The identification unit When the box image is detected, pixels having pixel values ​​according to a hue that belongs to a first hue range that is predetermined for the color of the number among the plurality of pixels that constitute the box image are converted to a predetermined first pixel value, and pixel values ​​of the remaining pixels are converted to a predetermined second pixel value, thereby converting the box image into a binary image consisting of pixels having the first pixel value and the second pixel value. Then, the box image converted into the binary image is applied as an input to the number identification model, and the first metric value consisting of a number is identified from the box image. The collection server according to claim 2 .

4. and a history information providing unit that, when a command requesting provision of metric value collection history information for a first period designated by an administrator is received from a pre-designated administrator terminal, extracts first metric value collection history information having date / time information belonging to the first period from among a plurality of pieces of metric value collection history information stored in the history database, and transmits the extracted first metric value collection history information to the administrator terminal. The collection server of claim 1 .

5. The history information providing unit a function storage unit in which a pseudo-random number generation function, which is shared in advance with the administrator terminal, is stored; an extracting unit that, when a command requesting provision of metric value collection history information for the first period designated by the administrator is received from the administrator terminal, checks the first metric value collection history information having date / time information belonging to the first period among a plurality of pieces of metric value collection history information stored in the history database, and extracts the first metric value collection history information from the history database; an encryption event generating unit that, when the first metric collection history information is extracted, encrypts the first metric collection history information and generates an encryption event for transmitting the first metric collection history information to the manager terminal; a random number generation unit that, when the encryption event occurs, applies a seed value as an input to the pseudo-random number generation function and repeatedly executes a process of generating random numbers n times (n is a natural number equal to or greater than 4), in which, in a first random number generation process, a randomly generated initial seed value is applied as an input to the pseudo-random number generation function, and from a second random number generation process onwards, a process of specifying the random number generated in the previous step as a seed value and applying it as an input to the pseudo-random number generation function is executed, thereby generating a total of n random numbers; a hash value generation unit that, when the n random numbers are sorted in ascending order and quartiles are calculated, selects random numbers from the n random numbers having a magnitude equal to or greater than the first quartile and equal to or less than the third quartile, sequentially concatenates the selected random numbers to generate concatenated values, and applies the concatenated values ​​as input to a preset hash function (the hash function is also stored in the administrator terminal) to generate hash values; and an information transmission unit that encrypts the first metric value collection history information based on the hash value, and then transmits the encrypted first metric value collection history information and the initial seed value to the administrator terminal; further comprising The administrator terminal When encrypted first metric collection history information and the initial seed value are received from the collection server, a step of applying the seed value as an input to the pseudo-random number generation function stored in the administrator terminal to generate random numbers is repeatedly executed n times. In the first random number generation step, the initial seed value is applied as an input to the pseudo-random number generation function. In the second random number generation step, the random number generated in the previous step is designated as a seed value and applied as an input to the pseudo-random number generation function. After the n random numbers are generated, the n random numbers are sorted in ascending order to calculate quartiles. Then, random numbers having magnitudes equal to or greater than the first quartile and equal to or less than the third quartile are selected from the n random numbers, and the selected random numbers are sequentially concatenated to generate concatenated values. The concatenated values ​​are applied as input to the hash function stored in the administrator terminal to generate the hash values. Then, the encrypted first metric collection history information is decrypted based on the hash values. The collection server according to claim 4 .

6. A method for operating an artificial intelligence-based collection server that collects measurement values ​​in conjunction with a camera arranged to photograph an area where measurement values ​​are displayed on a weighing device, comprising: generating a collection event for collecting metric values ​​at a predetermined collection cycle interval; a step of transmitting, when the collection event occurs in any cycle during the step of generating the collection event, a photographing request command to request the camera to photograph an area on the meter where the measurement value is displayed and transmit the photograph; When a captured image of the area of ​​the weighing scale where the measurement value is displayed is transferred from the camera in response to the image capture request command, a step of identifying a first measurement value consisting of a numerical value from the captured image based on a preset number identification model (the number identification model is an artificial intelligence-based model that has been machine-learned in advance to identify numbers present in the image and calculate an identification result therefrom); and generating date / time information of the time when the collection event occurred, generating metric collection history information consisting of the date / time information and the first metric value, and then storing the metric collection history information in a history database in which multiple pieces of metric collection history information are stored. A method for operating a collection server, comprising:

7. The identifying step includes: When the captured image is transferred from the camera, a rectangular box image for an object corresponding to an area where a metric value exists is detected from the captured image based on a preset object detection model (the object detection model is a model based on artificial intelligence that has been machine-learned in advance to detect an object corresponding to an area where a metric value exists in an image in the form of a rectangular box), and the box image is applied as an input to the digit identification model to identify a first metric value consisting of a digit from the box image.

7. A method for operating a collection server according to claim 6.

8. The identifying step includes: When the box image is detected, pixels having pixel values ​​according to a hue that belongs to a first hue range that is predetermined for the color of the number among the plurality of pixels that constitute the box image are converted to a predetermined first pixel value, and pixel values ​​of the remaining pixels are converted to a predetermined second pixel value, thereby converting the box image into a binary image consisting of pixels having the first pixel value and the second pixel value, and then applying the box image converted into a binary image to the number identification model as an input, and identifying the first metric value consisting of a numerical value from the box image. A method for operating a collection server according to claim 7.

9. and when a command to request provision of metric value collection history information for a first period designated by an administrator is received from a pre-designated administrator terminal, extracting first metric value collection history information having date / time information belonging to the first period from among a plurality of metric value collection history information stored in the history database, and transmitting the extracted first metric value collection history information to the administrator terminal.

7. A method for operating a collection server according to claim 6.

10. The step of transmitting to the manager terminal includes: a step of retaining a function storage unit in which a pseudo-random number generation function, which is shared in advance with the administrator terminal, is stored; When a command to request provision of metric value collection history information for the first period designated by the administrator is received from the administrator terminal, checking the first metric value collection history information having date / time information belonging to the first period among a plurality of metric value collection history information stored in the history database, and extracting the first metric value collection history information from the history database; generating an encryption event for encrypting the first metric collection history information and transmitting it to the administrator terminal when the first metric collection history information is extracted; When the encryption event occurs, a process of applying a seed value as an input to the pseudo-random number generation function and generating random numbers is repeatedly executed n times (n is a natural number equal to or greater than 4), in which in a first random number generation process, a randomly generated initial seed value is applied as an input to the pseudo-random number generation function, and from a second random number generation process onwards, a process of specifying the random number generated in the previous step as a seed value and applying it as an input to the pseudo-random number generation function is executed, thereby generating n random numbers in total; When the n random numbers are sorted in ascending order and quartiles are calculated, selecting random numbers from the n random numbers having a magnitude equal to or greater than the first quartile and equal to or less than the third quartile, sequentially concatenating the selected random numbers to generate concatenated values, and applying the concatenated values ​​as input to a preset hash function (the hash function is also stored in the administrator terminal) to generate a hash value; and and further comprising: encrypting the first metric collection history information based on the hash value, and then transmitting the encrypted first metric collection history information and the initial seed value to the administrator terminal; The administrator terminal When the encrypted first metric collection history information and the initial seed value are received from the collection server, a step of applying the seed value as an input to the pseudo-random number generation function stored in the administrator terminal to generate random numbers is repeatedly executed n times. In the first random number generation step, the initial seed value is applied as an input to the pseudo-random number generation function. In the second random number generation step, the random number generated in the previous step is designated as a seed value and applied as an input to the pseudo-random number generation function. After the n random numbers are generated, the n random numbers are sorted in ascending order to calculate quartiles. Then, random numbers having magnitudes equal to or greater than the first quartile and equal to or less than the third quartile are selected from the n random numbers, and the selected random numbers are sequentially concatenated to generate concatenated values. The concatenated values ​​are applied as input to the hash function stored in the administrator terminal to generate the hash values. Then, the encrypted first metric collection history information is decrypted based on the hash values.

10. A method for operating a collection server according to claim 9.

11. A computer program for executing the collection server operation method according to any one of claims 6 to 10 through connection with a computer is recorded. A computer-readable recording medium.

12. A method for operating a collection server according to any one of claims 6 to 10, stored in a storage medium for executing the method through connection with a computer. A computer program characterized by:

Images