Image processing device, image processing method, and recording medium

Abstract

Provided is an image processing device, wherein an image acquiring means acquires a camera image that contains at least one analog meter and an object that differs from the analog meter, and also acquires a master image captured at a position closest to a position where the analog meter contained in the camera image has been captured. An identification means identifies, on the basis of processing results obtained by carrying out feature-point matching on the entire area of the camera image and the entire area of the master image, a designated analog meter contained in the camera image and the master image. An indicated-value reading means performs a process for reading an indicated value on the designated analog meter.

Description

Image processing device, image processing method, and recording medium

[0001] The present disclosure relates to a technique that can be used to read the indicated value of an analog meter included in an image.

[0002] A technique has been proposed that can be used to read the indicated value of an analog meter included in an image.

[0003] Specifically, for example, Patent Document 1 discloses a method of identifying the configuration (type) of an analog meter from an image including a circular area corresponding to the analog meter, and reading the value measured by the analog meter based on the identified configuration (type).

[0004] Japanese Patent Application Laid-Open No. 2023-29227

[0005] However, the technology disclosed in Patent Document 1 has the problem that, for example, if information for identifying the configuration of an analog meter cannot be obtained from the image, it may not be possible to read the indicated value of the analog meter.

[0006] One object of the present disclosure is to provide an image processing device that can read the indicated value of an analog meter even when information for identifying the configuration of the analog meter cannot be obtained from the image.

[0007] In one aspect of the present disclosure, an image processing device has an image acquisition means for acquiring a camera image including at least one analog meter and an object other than the analog meter, and acquiring a master image captured at a position closest to the position where the analog meter included in the camera image was photographed, an identification means for identifying a specified analog meter included in the camera image and the master image based on processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image, and an indication value reading means for performing processing to read the indication value of the specified analog meter.

[0008] In another aspect of the present disclosure, an image processing method is an image processing method executed by a computer, which acquires a camera image including at least one analog meter and an object other than the analog meter, acquires a master image captured at a position closest to the position where the analog meter included in the camera image was captured, and identifies a specified analog meter included in the camera image and the master image based on the processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image, and performs processing to read the indicated value of the specified analog meter.

[0009] In yet another aspect of the present disclosure, a recording medium records a program that causes a computer to acquire a camera image including at least one analog meter and an object other than the analog meter, acquire a master image captured at a position closest to the position where the analog meter included in the camera image was captured, and identify a specified analog meter included in the camera image and the master image based on the processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image, and read the indicated value of the specified analog meter.

[0010] According to the present disclosure, even when information for identifying the configuration of an analog meter cannot be obtained from an image, the indicated value of the analog meter can be read.

[0011] 6 is a diagram showing a schematic configuration of an image processing system including an image processing device according to the present disclosure. FIG. 6 is a block diagram showing an example of a hardware configuration of an image processing device according to the present disclosure. FIG. 6 is a block diagram showing an example of a functional configuration of an image processing device according to the present disclosure. FIG. 6 is a diagram showing an example of a camera image used in processing of an image processing device according to the present disclosure. FIG. 6 is a diagram showing an example of a master image used in processing of an image processing device according to the present disclosure. FIG. 6 is a diagram for explaining processing performed on the corrected camera image of FIG. 6. FIG. 6 is a diagram showing an example of an image cut out from the corrected camera image of FIG. 6. FIG. 6 is a diagram for explaining an example of processing performed by an image processing device according to the present disclosure. A flowchart showing an example of processing performed in an image processing device according to the present disclosure. A block diagram showing another example of the functional configuration of an image processing device according to the present disclosure. A flowchart showing another example of processing performed in an image processing device according to the present disclosure.

[0012] Hereinafter, preferred embodiments of the present disclosure will be described with reference to the drawings.

[0013] <First embodiment> [System configuration] Fig. 1 is a diagram showing a schematic configuration of an image processing system including an image processing device according to the present disclosure. As shown in Fig. 1, the image processing system 1 has a camera 10 and an image processing device 100. The camera 10 and the image processing device 100 are configured to be able to communicate with each other via a network NW.

[0014] The camera 10 is mounted on a robot that patrols a predetermined facility such as a factory. The camera 10 captures images by photographing objects such as analog meters installed within the facility. The images captured by the camera 10 (hereinafter also referred to as camera images) are output to the image processing device 100 via the network NW. In the following description, unless otherwise specified, it is assumed that the camera image includes at least one circular analog meter installed in the predetermined facility and an object other than the analog meter that is present in the facility.

[0015] The image processing device 100 can acquire a camera image including at least one analog meter and an object other than the analog meter, and can acquire a master image captured at a position closest to the position where the analog meter in the camera image was captured. The image processing device 100 can identify a specific analog meter included in the camera image and the master image based on the results of a feature point matching process performed on the entire camera image and the entire master image. The image processing device 100 can acquire a corrected camera image by performing a correction process in accordance with the results of the feature point matching process, such as adjusting the position of the specific analog meter in the camera image to match the position of the specific analog meter in the master image. The image processing device 100 can also acquire a cropped image by cropping an area including the specific analog meter from the corrected camera image. The image processing device 100 can also perform a process to read the indicated value of the specific analog meter in the cropped image. Details of each process performed by the image processing device 100 will be described later. In this embodiment, the term "indicated value of an analog meter" can be rephrased as "indicated value indicated by the meter needle of an analog meter."

[0016] [Hardware Configuration] Fig. 2 is a block diagram showing an example of the hardware configuration of an image processing device according to the present disclosure. As shown in Fig. 2, the image processing device 100 includes an interface (IF) 111, a processor 112, a memory 113, a recording medium 114, and a database (DB) 115.

[0017] The IF 111 inputs and outputs data to and from an external device. For example, a camera image is input to the image processing device 100 through the IF 111.

[0018] The processor 112 is a computer such as a CPU (Central Processing Unit), and executes a prepared program to control the entire image processing device 100. Specifically, the processor 112 performs processing related to reading the indicated value of an analog meter included in a camera image, for example.

[0019] The memory 113 is configured by a ROM (Read Only Memory), a RAM (Random Access Memory), etc. The memory 113 is also used as a working memory while the processor 112 is executing various processes.

[0020] The recording medium 114 is a non-volatile, non-transitory recording medium such as a disk-shaped recording medium or a semiconductor memory, and is configured to be detachable from the image processing device 100. The recording medium 114 records various programs to be executed by the processor 112. When the image processing device 100 executes various processes, the programs recorded on the recording medium 114 are loaded into the memory 113 and executed by the processor 112.

[0021] The DB 115 stores information input via the IF 111, processing results obtained by processing by the processor 112, etc. Specifically, the DB 115 stores, for example, the reading results of the indicated value of an analog meter included in a camera image.

[0022] [Functional Configuration] Fig. 3 is a block diagram showing an example of the functional configuration of an image processing device according to the present disclosure. As shown in Fig. 3, the image processing device 100 includes an image storage unit 11, an image acquisition unit 12, a correction processing unit 13, a cutout processing unit 14, an area extraction unit 15, and an indication value reading unit 16.

[0023] A plurality of master images MG obtained by photographing objects such as analog meters installed in a predetermined facility are stored in the image storage unit 11. In the following description, unless otherwise specified, it is assumed that each master image MG includes at least one circular analog meter installed in the predetermined facility and an object other than the analog meter that is present in the predetermined facility.

[0024] The image acquisition unit 12 functions as an image acquisition means. The image acquisition unit 12 also acquires from the image storage unit 11 a master image MG captured at a position closest to the position at which the analog meter included in the camera image CG obtained via the network NW was captured, and outputs the camera image CG and the master image MG to the correction processing unit 13.

[0025] The correction processing unit 13 functions as an identification unit and a correction processing unit. The correction processing unit 13 also has a feature extraction model 13A that is trained to extract features of an input image and output a feature map. The feature map can be treated, for example, as data indicating the magnitude of the feature amount at each pixel position included in the input image. The correction processing unit 13 also inputs a camera image CG obtained from the image acquisition unit 12 into the feature extraction model 13A to acquire a feature map TC corresponding to the entire camera image CG. The correction processing unit 13 also inputs a master image MG obtained from the image acquisition unit 12 into the feature extraction model 13A to acquire a feature map TM corresponding to the entire master image MG.

[0026] The correction processing unit 13 also performs a feature point matching process to identify pixel positions where the feature amounts in the feature map TC and the feature amounts in the feature map TM are similar. The correction processing unit 13 also identifies a predetermined analog meter included in the camera image CG and the master image MG based on the processing results of the feature point matching process described above. That is, the correction processing unit 13 can identify a predetermined analog meter included in the camera image CG and the master image MG based on the processing results obtained by performing the feature point matching process on the entire area of ​​the camera image CG and the entire area of ​​the master image MG.

[0027] Furthermore, the correction processing unit 13 performs a correction process according to the result of the feature point matching process, for example, a process of correcting the position and shape of a predetermined analog meter included in the camera image CG so that the position and shape of the predetermined analog meter best match the position and shape of the predetermined analog meter included in the master image MG. The correction processing unit 13 also outputs to the cutout processing unit 14 the master image MG used in the feature point matching process and a corrected camera image AG obtained by applying the correction process to the camera image CG.

[0028] The correction processing unit 13 may perform, as a correction process according to the results of the feature point matching process, at least a process of adjusting the position of a predetermined analog meter in the camera image CG so that the position of the predetermined analog meter matches the position of the predetermined analog meter in the master image MG. The correction process may also include, for example, a process of deforming and / or moving at least a portion of the camera image CG according to the results of the feature point matching process so that the position and shape of the predetermined analog meter in the camera image CG best matches the position and shape of the predetermined analog meter in the master image MG. The correction process may also include a process related to coordinate transformation of a two-dimensional image. According to the present disclosure, the correction processing unit 13 may be configured to perform a process similar to the correction process described above even if the predetermined analog meter included in the camera image CM and the master image MG has a shape other than a circle.

[0029] The cutout processing unit 14 functions as a cutout unit. The cutout processing unit 14 performs a cutout process using the master image MG and the corrected camera image AG obtained from the correction processing unit 13, thereby obtaining an image (hereinafter also referred to as a cutout image KG) in which a predetermined analog meter is cut out from the corrected camera image AG. Specifically, the cutout processing unit 14 obtains the cutout image KG by, for example, cutting out from the corrected camera image AG an area identical to an area including the predetermined analog meter in the master image MG. The cutout processing unit 14 outputs the master image MG used in the cutout process and the cutout image KG obtained as a result of the cutout process to the area extraction unit 15.

[0030] The region extraction unit 15 functions as a region extraction means. The region extraction unit 15 has a region extraction model 15A that has been trained to output an extraction result indicating a region corresponding to the meter needle of a predetermined analog meter in response to an input of an image including an analog meter. The region extraction unit 15 inputs the cutout image KG obtained from the cutout processing unit 14 to the region extraction model 15A, thereby obtaining an extraction result ER indicating a region in the cutout image KG corresponding to the meter needle of a predetermined analog meter. The region extraction unit 15 outputs the master image MG and cutout image KG obtained from the cutout processing unit 14, as well as the extraction result ER corresponding to the cutout image KG, to the indicated value reading unit 16.

[0031] According to the present disclosure, when training the region extraction model 15A, it is desirable to use, as training data, data in which, for example, an image including an analog meter is labeled with a region corresponding to the meter needle. Also, according to the present disclosure, when training the region extraction model 15A, in addition to the above-mentioned data, data in which, for example, an image including a clock is labeled with a region corresponding to the hour hand, minute hand, and second hand can be used as training data. Also, according to the present disclosure, for example, a process of adjusting the brightness of the image included in the training data may be performed before training the region extraction model 15A. Also, according to the present disclosure, the region extraction unit 15 may, for example, perform a process of adjusting the brightness of the cutout image KG before inputting the cutout image KG to the region extraction model 15A.

[0032] The indicated value reading unit 16 functions as an indicated value reading means. The indicated value reading unit 16 performs a reading process using the master image MG, the cutout image KG, and the extraction result ER obtained from the area extraction unit 15, thereby reading the indicated value SV indicated by the meter needle of a predetermined analog meter in the cutout image KG. That is, the indicated value reading unit 16 can perform a process for reading the indicated value SV of the predetermined analog meter identified by the correction processing unit 13. Specific examples of the aforementioned reading process will be described later. The indicated value reading unit 16 also outputs a reading result YR of the indicated value SV obtained by the aforementioned reading process to the outside. The reading result YR can be used as a reading result of the indicated value indicated by the meter needle of a predetermined analog meter in the camera image CG.

[0033] [Specific Example] Next, a specific example of the processing etc. performed in the present disclosure will be described.

[0034] Fig. 4 is a diagram showing an example of a camera image used for processing by the image processing device according to the present disclosure. The camera 10 captures an image of an object in a predetermined facility, thereby acquiring, for example, a camera image CG1 as shown in Fig. 4. The camera image CG1 includes an analog meter CM and an object CB, such as a pipe, present around the analog meter CM. Note that, for ease of illustration, the text on the meter face of the analog meter CM is omitted in Fig. 4. Furthermore, the analog meter CM in Fig. 4 has a vertically elongated oval shape, for example, because it was captured at an angle other than the front.

[0035] The camera 10 acquires shooting position information CS indicating the position where an analog meter CM was captured within a specified facility, and associates the acquired shooting position information CS with the camera image CG1. The camera 10 also outputs the camera image CG1 associated with the shooting position information CS to the network NW. The shooting position information CS can be treated as information indicating the position where the camera image CG1 was captured within the specified facility. The shooting position information CS can also include, for example, information obtained from a Global Positioning System (GPS) and / or information obtained from a Light Detection and Ranging (LiDAR).

[0036] Each of the master images MG stored in the image storage unit 11 is associated with photographing position information MS, which includes the same type of information as the photographing position information CS, as information indicating the position at which a specific analog meter was photographed within a specific facility. The photographing position information MS can be treated as information indicating the position at which the master image MG was acquired within the specific facility.

[0037] The shape of the analog meter included in each of the multiple master images MG is corrected in advance so that it has the same shape as the analog meter when viewed from the front. In this specific example, unless otherwise specified, it is assumed that the shape of the analog meter included in each of the multiple master images MG has been corrected to a perfect circle. The process of correcting the shape of the analog meter in each of the multiple master images MG to a perfect circle may be performed manually or according to a predetermined algorithm.

[0038] Meter coordinate information MZ, which is information relating to the position of a predetermined analog meter in the image, is associated with each of the multiple master images MG stored in the image storage unit 11. The meter coordinate information MZ includes meter area information MJ1, rotation center coordinate information MJ2, minimum value coordinate information MJ3, and maximum value coordinate information MJ4, which are information relating to the position of the predetermined analog meter corrected to a perfect circle shape in the master image MG.

[0039] The meter area information MJ1 includes coordinate information used in the processing of the correction processing unit 13 and the cutout processing unit 14. Specifically, the meter area information MJ1 includes, for example, coordinates capable of identifying a frame of a predetermined shape that indicates an area corresponding to a predetermined analog meter in the master image MG. In this specific example, unless otherwise specified, it is assumed that the meter area information MJ1 includes coordinates capable of identifying a rectangular frame that indicates an area corresponding to a predetermined analog meter in the master image MG. The rotation center coordinate information MJ2 includes coordinates that indicate the rotation center of the meter needle of the predetermined analog meter in the master image MG. The minimum value coordinate information MJ3 includes coordinates that indicate the position of the minimum value (hereinafter also referred to as the minimum indicated value) of the indicated value written on the meter face of the predetermined analog meter in the master image MG. The maximum value coordinate information MJ4 includes coordinates that indicate the position of the maximum value (hereinafter also referred to as the maximum indicated value) of the indicated value written on the meter face of the predetermined analog meter in the master image MG.

[0040] Meter value information MV, which is information relating to the meter value of a predetermined analog meter in the image, is associated with each of the multiple master images MG stored in the image storage unit 11. The meter value information MV includes information indicating the minimum indicated value VS and the maximum indicated value VD of the predetermined analog meter in the master image MG.

[0041] FIG. 5 is a diagram illustrating an example of a master image used in processing by the image processing device according to the present disclosure. The image acquisition unit 12 acquires one master image MG1 from among multiple master images MG stored in the image storage unit 11, the master image MG1 being associated with shooting position information MS indicating a position closest to the shooting position information CS of the camera image CG1 acquired via the network NW. Through this processing, the image acquisition unit 12 can acquire, for example, the master image MG1 shown in FIG. 5 . The master image MG1 includes an analog meter MM and an object MB, such as a pipe, present around the analog meter MM. Note that, for ease of illustration, the character string on the meter face of the analog meter MM is omitted in FIG. 5 . The image acquisition unit 12 outputs the camera image CG1 and the master image MG1 to the correction processing unit 13.

[0042] The correction processing unit 13 inputs the camera image CG1 obtained from the image acquisition unit 12 into the feature extraction model 13A to obtain a feature map TC1 corresponding to the camera image CG1. The correction processing unit 13 also inputs the master image MG1 obtained from the image acquisition unit 12 into the feature extraction model 13A to obtain a feature map TM1 corresponding to the master image MG1.

[0043] The correction processing unit 13 also performs a feature point matching process to identify pixel positions where the feature amounts in the feature map TC1 and the feature amounts in the feature map TM1 are similar. The correction processing unit 13 also references the results of the feature point matching process and the meter region information MJ1 included in the meter coordinate information MZ to identify pixel groups having feature amounts similar to the feature amounts of the pixel group PG1 that constitute the analog meter MM in the master image MG1 as the pixel group PG2 that constitutes the analog meter CM in the camera image CG1. That is, the correction processing unit 13 can identify the analog meter CM in the camera image CG1 and the analog meter MM in the master image MG1 as the same analog meter based on the processing results obtained by performing the feature point matching process on the entire area of ​​the camera image CG1 and the entire area of ​​the master image MG1.

[0044] Furthermore, as a correction process according to the results of identifying the pixel groups PG1 and PG2, the correction processing unit 13 performs a process of deforming and / or moving at least a portion of the camera image CG1 so that the position and shape of the analog meter CM in the camera image CG1 best matches the position and shape of the analog meter MM in the master image MG1. The correction processing unit 13 also outputs to the cutout processing unit 14 the master image MG1 used in the feature point matching process and a corrected camera image AG1 obtained by applying the correction process to the camera image CG1.

[0045] Fig. 6 is a diagram showing an example of a corrected camera image obtained by processing by the image processing device according to the present disclosure. According to the correction processing described above, the correction processing unit 13 can obtain, for example, a corrected camera image AG1 as shown in Fig. 6. The corrected camera image AG1 includes an analog meter AM and an object AB, such as a pipe, present around the analog meter AM. Note that, for ease of illustration, the character strings on the meter face of the analog meter AM are omitted in Fig. 6.

[0046] The analog meter AM represents a corrected analog meter in which the shape of the analog meter CM has been corrected to a perfect circle and the position of the analog meter CM has been corrected to a position corresponding to the meter area information MJ1. The analog meter AM also includes a meter needle having the same shape as the analog meter CM. The object AB represents an object common to each object CB included in the camera image CG1 and each object MB included in the master image MG1. According to this specific example, the correction processing unit 13 may fill in areas of the corrected camera image AG1 that are not included in the angle of view of the camera image CG1 with a predetermined color, such as black.

[0047] According to the processing described above, when the shape of the analog meter MM included in the master image MG1 is the same as the shape of the analog meter MM when viewed from the front, the correction processing unit 13 performs processing to deform the shape of the analog meter CM in the camera image CG1 so that it matches the shape of the analog meter MM in the master image MG1, as correction processing in accordance with the processing result of the feature point matching processing. Also, according to the processing described above, when the shape of the analog meter MM included in the master image MG1 is a perfect circle, the correction processing unit 13 performs processing to deform the shape of the analog meter CM in the camera image CM1 into a perfect circle, as correction processing in accordance with the processing result of the feature point matching processing.

[0048] Fig. 7 is a diagram for explaining the processing performed on the corrected camera image of Fig. 6. The cutout processing unit 14 acquires a rectangular frame RW corresponding to the position of the analog meter MM in the master image MG1 by referring to the meter region information MJ1 included in the meter coordinate information MZ of the master image MG1 obtained from the correction processing unit 13 (see Fig. 7).

[0049] FIG. 8 shows an example of an image cut out from the corrected camera image of FIG. 6 . In this example, the correction processing performed by the correction processing unit 13 adjusts the position of the analog meter AM in the corrected camera image AG1 to best match the position of the analog meter MM in the master image MG1. Therefore, the cutout processing unit 14 cuts out an area corresponding to the rectangular frame RW from the corrected camera image AG1 obtained from the correction processing unit 13, thereby obtaining a cutout image KG1 as shown in FIG. 8 . That is, the cutout processing unit 14 performs cutout processing using the master image MG1 and the corrected camera image AG1 obtained from the correction processing unit 13, thereby obtaining an image in which the analog meter AM has been cut out from the corrected camera image AG1. The cutout processing unit 14 also outputs the master image MG1 and the cutout image KG1 to the region extraction unit 15.

[0050] The region extraction unit 15 inputs the cutout image KG1 obtained from the cutout processing unit 14 into the region extraction model 15A, thereby obtaining an extraction result ER1 that indicates the region in the cutout image KG1 that corresponds to the meter needle ND of the analog meter AM. Through this processing, the region extraction unit 15 can obtain, for example, the hatched region in the cutout image KG1 in Figure 8 as the extraction result ER1. The region extraction unit 15 also outputs the master image MG1 and cutout image KG1 obtained from the cutout processing unit 14, and the extraction result ER1 corresponding to the cutout image KG1, to the indicated value reading unit 16.

[0051] The indication value reading unit 16 acquires the coordinate Po corresponding to the rotation center of the meter needle ND of the analog meter AM in the cut-out image KG1 obtained from the area extraction unit 15 based on the rotation center coordinate information MJ2 contained in the meter coordinate information MZ of the master image MG obtained from the area extraction unit 15.

[0052] The indicated value reading unit 16 acquires the coordinate Pa corresponding to the position of the minimum indicated value of the analog meter AM in the cutout image KG1 obtained from the area extraction unit 15, based on the minimum value coordinate information MJ3 included in the meter coordinate information MZ of the master image MG obtained from the area extraction unit 15. According to this processing, the indicated value reading unit 16 can acquire, for example, the coordinate Pa corresponding to the position of "0.00" on the analog meter AM in the cutout image KG1 of FIG.

[0053] The indicated value reading unit 16 acquires the coordinate Pb corresponding to the position of the maximum indicated value of the analog meter AM in the cutout image KG1 obtained from the area extraction unit 15, based on the maximum value coordinate information MJ4 included in the meter coordinate information MZ of the master image MG obtained from the area extraction unit 15. According to this processing, the indicated value reading unit 16 can acquire, for example, the coordinate Pb corresponding to the position of "0.50" on the analog meter AM in the cutout image KG1 of FIG.

[0054] The indicated value reading unit 16 uses the coordinate Po and the extraction result ER1 to obtain the coordinate Pc corresponding to the tip of the meter needle ND of the analog meter AM in the cutout image KG1.

[0055] Here, a specific example of the process for obtaining the coordinate Pc will be described.

[0056] The indication value reading unit 16 extracts a plurality of line segment candidates LN constituting the region corresponding to the extraction result ER1 in the cutout image KG1 by, for example, applying a Hough transform to the region. The indication value reading unit 16 also selects the longest line segment candidate LN1 that passes through the coordinate Po from the plurality of line segment candidates LN, and acquires the selected line segment candidate LN1 as the line segment representing the meter needle ND of the analog meter AM in the cutout image KG1. The indication value reading unit 16 also acquires the coordinate of the farthest of the two endpoints of the line segment candidate LN1 from the coordinate Po as the coordinate Pc corresponding to the tip of the meter needle ND of the analog meter AM in the cutout image KG1.

[0057] According to the above-described process, the indication value reading unit 16 can obtain the coordinate Pc based on the extraction result ER1.

[0058] The indication value reading unit 16 identifies the minimum indication value VS1 and the maximum indication value VD1 of the analog meter AM by referring to the meter value information MV of the master image MG1. Through this processing, the indication value reading unit 16 can identify that the minimum indication value VS1 of the analog meter AM is "0.00" and the maximum indication value VD1 of the analog meter AM is "0.50."

[0059] The indication value reading unit 16 reads the indication value SV1 indicated by the meter needle ND of the analog meter AM in the cutout image KG1 by performing a reading process using the coordinates Po, Pa, Pb and Pc, the minimum indication value VS1 and the maximum indication value VD1.

[0060] A specific example of the reading process of the indication value SV1 will now be described with reference to Fig. 9. Fig. 9 is a diagram for explaining an example of the process performed by the image processing apparatus according to the present disclosure.

[0061] The indication value reading unit 16 sets the coordinate Po as the origin of the Cartesian coordinate system, and then obtains a vector Va pointing from the coordinate Po to the coordinate Pa, a vector Vb pointing from the coordinate Po to the coordinate Pb, and a vector Vc pointing from the coordinate Po to the coordinate Pc (see FIG. 9 ). The indication value reading unit 16 also calculates an angle θ1 (see FIG. 9 ) formed by the vector Va and the vector Vb by applying the vector Va and the vector Vb to the following equation (1):

[0062]

[0063] That is, the indicated value reading unit 16 can calculate the angle θ1 as an angle corresponding to the movable range of the meter needle ND of the analog meter AM.

[0064] The indication value reading unit 16 calculates the angle θ2 (see FIG. 9) formed by the vector Va and the vector Vc by applying the vector Va and the vector Vc to the following equation (2).

[0065]

[0066] The indication value reading unit 16 calculates the change UV in the indication value of the analog meter AM corresponding to an angle of 1° by applying the minimum indication value VS1, the maximum indication value VD1, and the angle θ1 to the following formula (3).

[0067]

[0068] The indication value reading unit 16 calculates the indication value SV1 indicated by the meter needle ND of the analog meter AM by applying the minimum indication value VS1, the amount of change UV, and the angle θ2 to the following equation (4).

[0069]

[0070] The indicated value reading unit 16 outputs the calculation result of the indicated value SV1 obtained by the above-described process to the outside as a reading result YR1. The reading result YR1 can be used as a reading result of the indicated value indicated by the meter needle of the analog meter CM in the camera image CG1.

[0071] According to this specific example, for example, when the analog meter CM in the camera image MG1 and the analog meter MM in the master image MG1 do not have a perfect circular shape, the correction processing unit 13 does not need to perform processing to deform the shape of the analog meter CM so as to match the shape of the analog meter MM. In other words, when the analog meter CM in the camera image MG1 and the analog meter MM in the master image MG1 do not have a perfect circular shape, the correction processing unit 13 only needs to perform processing to adjust the position of the analog meter CM in the camera image MG1 so as to match the position of the analog meter MM in the master image MG1, as a correction processing in accordance with the processing result of the feature point matching processing.

[0072] According to this specific example, for example, when the analog meter CM in the camera image MG1 and the analog meter MM in the master image MG1 have a perfect circular shape, the correction processing unit 13 does not need to perform processing to deform the shape of the analog meter CM so as to match the shape of the analog meter MM. In other words, when the analog meter CM in the camera image MG1 and the analog meter MM in the master image MG1 have a perfect circular shape, the correction processing unit 13 only needs to perform processing to adjust the position of the analog meter CM in the camera image MG1 so as to match the position of the analog meter MM in the master image MG1, as a correction processing in accordance with the processing result of the feature point matching processing.

[0073] According to this specific example, the image processing device 100 performs feature point matching processing on the entire area of ​​the camera image CG1 and the entire area of ​​the master image MG1. Furthermore, for example, when the camera image CM1 includes multiple analog meters, the image processing device 100 can identify one analog meter CM among the multiple analog meters as the same analog meter as the analog meter MM based on the processing results of the feature point matching processing. Therefore, even when the camera image CG1 includes multiple analog meters, the image processing device 100 can read the indicated value indicated by the meter needle of one analog meter CM among the multiple analog meters.

[0074] [Processing Flow] Next, a description will be given of the flow of processing performed in the image processing device 100. Fig. 10 is a flowchart showing an example of processing performed in the image processing device according to the present disclosure.

[0075] First, the image processing device 100 acquires from the image storage unit 11 a master image MG captured at a position closest to the position where the analog meter included in the camera image CG acquired via the network NW was captured (step S11).

[0076] Next, the image processing device 100 inputs the camera image CG and the master image MG obtained in step S11 into the feature extraction model 13A, thereby obtaining a feature map TC corresponding to the camera image CG and a feature map TM corresponding to the master image MG (step S12).

[0077] Next, the image processing device 100 performs a feature point matching process to identify pixel positions where the feature amounts in the feature map TC obtained in step S12 are similar to the feature amounts in the feature map TM obtained in step S12 (step S13).

[0078] Next, the image processing device 100 performs a correction process on the camera image CM obtained in step S11 in accordance with the processing result of step S13, thereby obtaining a corrected camera image AG (step S14). By performing the correction process of step S14, the image processing device 100 can obtain a corrected camera image AG in which, for example, the position of a predetermined analog meter in the camera image CG has been adjusted to match the position of the predetermined analog meter in the master image MG.

[0079] Next, the image processing device 100 performs a process of cutting out an area from the corrected camera image AG obtained in step S13 that includes an analog meter that is identical to a specified analog meter in the master image MG obtained in step S11, thereby obtaining a cut-out image KG (step S15).

[0080] Next, the image processing device 100 inputs the cut-out image KG obtained in step S15 into the area extraction model 15A, thereby obtaining an extraction result ER indicating the area in the cut-out image KG corresponding to the meter needle of a specified analog meter (step S16).

[0081] Next, the image processing device 100 reads the indication value SV indicated by the meter needle of the specified analog meter by processing using the coordinate information for the specified analog meter in the cut-out image KG obtained in step S15, the minimum indication value VS and the maximum indication value VD of the specified analog meter, and the extraction result ER obtained in step S16 (step S17).

[0082] As described above, according to this embodiment, by performing feature point matching processing on the entire area of ​​a camera image and the entire area of ​​a master image, it is possible to identify a specific analog meter included in the camera image and the master image. Therefore, according to this embodiment, even if a camera image is acquired including an analog meter whose meter face does not have an identifier such as a two-dimensional code, it is possible to identify the analog meter and read its indicated value. Furthermore, according to this embodiment, even if a camera image is acquired in which the meter face of the analog meter is unclear, it is possible to identify the analog meter and read its indicated value. Furthermore, according to this embodiment, even if information for identifying the configuration of the analog meter cannot be obtained from the image due to, for example, dirt and / or aging on the meter face of the analog meter, it is possible to identify the analog meter and read its indicated value. Furthermore, according to this embodiment, even if information for identifying the configuration of the analog meter cannot be obtained from the image due to, for example, an inappropriate position, size, or tilt of the analog meter in the camera image, it is possible to identify the analog meter and read its indicated value. Furthermore, according to this embodiment, even if an object similar to a specific analog meter is included in a camera image, the specific analog meter can be identified and the indicated value can be read.

[0083] As described above, according to this embodiment, the indicated value of an analog meter can be read after the shape of the analog meter included in the camera image has been corrected to a perfect circle. Therefore, according to this embodiment, even if the camera image includes a distorted analog meter due to, for example, being photographed from an angle other than the front, the accuracy of reading the indicated value of the analog meter can be ensured.

[0084] Second Embodiment FIG. 11 is a block diagram showing another example of the functional configuration of an image processing device according to the present disclosure.

[0085] The image processing device 500 has the same hardware configuration as the image processing device 100. The image processing device 500 also has an image acquisition unit 511, an identification unit 512, and an indication value reading unit 513.

[0086] The image acquisition means 511 can be realized, for example, by using the function of the image acquisition unit 12. The identification means 512 can be realized, for example, by using the function of the correction processing unit 13. The indication value reading means 513 can be realized, for example, by using the function of the indication value reading unit 16.

[0087] FIG. 12 is a flowchart showing another example of processing performed in the image processing device according to the present disclosure.

[0088] The image acquisition means 511 acquires a camera image including at least one analog meter and an object other than the analog meter, and also acquires a master image taken at a position closest to the position where the analog meter included in the camera image was photographed (step S51).

[0089] The identification means 512 identifies a predetermined analog meter contained in the camera image and the master image based on the processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image (step S52).

[0090] The indicated value reading means 513 performs processing to read the indicated value of a predetermined analog meter (step S53).

[0091] According to this embodiment, even if information for identifying the configuration of an analog meter cannot be obtained from an image, the indicated value of the analog meter can be read.

[0092] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes.

[0093] (Supplementary Note 1) An image processing device having: an image acquisition means for acquiring a camera image including at least one analog meter and an object other than the analog meter, and acquiring a master image captured at a position closest to the position at which the analog meter included in the camera image was captured; an identification means for identifying a specified analog meter included in the camera image and the master image based on a processing result obtained by performing a feature point matching process on the entire area of ​​the camera image and the entire area of ​​the master image; and an indication value reading means for performing processing to read the indication value of the specified analog meter.

[0094] (Supplementary Note 2) The image processing device of Supplementary Note 1 further comprises: a correction processing means for performing, as a correction process according to the processing result of the feature point matching process, a process for adjusting the position of the specified analog meter in the camera image so that it matches the position of the specified analog meter in the master image, thereby acquiring a corrected camera image; and a cutting means for performing a process for cutting out an area including the specified analog meter from the corrected camera image, thereby acquiring a cutout image, wherein the indication value reading means performs a process for reading the indication value of the specified analog meter in the cutout image.

[0095] (Supplementary Note 3) The image processing device of Supplementary Note 1, wherein the identification means, when the camera image includes multiple analog meters, identifies the predetermined analog meter from among the multiple analog meters based on a processing result obtained by performing the feature point matching process.

[0096] (Appendix 4) The image processing device of Appendix 2, wherein the correction processing means further performs, as the correction processing, a process of deforming the shape of the specified analog meter in the camera image so that it matches the shape of the specified analog meter in the master image when the shape of the specified analog meter included in the master image is the same as the shape of the specified analog meter when viewed from the front.

[0097] (Supplementary Note 5) The image processing device of Supplementary Note 4 further comprises an area extraction means for inputting the cutout image into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to a meter needle of the specified analog meter, wherein the indicated value reading means reads the indicated value of the specified analog meter in the cutout image by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

[0098] (Appendix 6) The image processing device of Appendix 2, wherein the correction processing means further performs processing to deform the shape of the specified analog meter in the camera image into a perfect circle as the correction processing when the shape of the specified analog meter included in the master image has a perfect circle shape.

[0099] (Supplementary Note 7) The image processing device of Supplementary Note 6 further comprises an area extraction means for inputting the cutout image into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to a meter needle of the specified analog meter, wherein the indicated value reading means reads the indicated value of the specified analog meter in the cutout image by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

[0100] (Supplementary Note 8) The image processing device according to Supplementary Note 1, wherein the camera image and the master image are linked with information relating to a position within a predetermined facility where the image was acquired.

[0101] (Supplementary Note 9) An information processing method executed by a computer, comprising: acquiring a camera image including at least one analog meter and an object other than the analog meter; acquiring a master image captured at a position closest to the position at which the analog meter included in the camera image was captured; performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image, based on the processing results obtained, identifying a specified analog meter included in the camera image and the master image; and performing processing to read the indicated value of the specified analog meter.

[0102] (Supplementary Note 10) The image processing method of Supplementary Note 9 includes, as a correction process according to the processing result of the feature point matching process, performing a process of adjusting the position of the specified analog meter in the camera image so that it matches the position of the specified analog meter in the master image, thereby obtaining a corrected camera image; performing a process of cutting out an area including the specified analog meter from the corrected camera image, thereby obtaining a cutout image; and performing a process to read the indicated value of the specified analog meter in the cutout image.

[0103] (Appendix 11) The image processing method of Appendix 10, further comprising, as the correction process, a process of deforming the shape of the specified analog meter in the camera image so that it matches the shape of the specified analog meter in the master image, when the shape of the specified analog meter included in the master image is the same as the shape of the specified analog meter when viewed from the front.

[0104] (Supplementary Note 12) The image processing method of Supplementary Note 11, in which the cutout image is input into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, and the indicated value of the specified analog meter in the cutout image is read by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

[0105] (Supplementary Note 13) The image processing method of Supplementary Note 10, further comprising, as the correction process, performing a process of deforming the shape of the specified analog meter in the camera image into a perfect circle when the shape of the specified analog meter included in the master image is a perfect circle.

[0106] (Supplementary Note 14) The image processing method of Supplementary Note 13, in which the cutout image is input into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, and the indicated value of the specified analog meter in the cutout image is read by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

[0107] (Supplementary Note 15) A recording medium having recorded thereon a program that causes a computer to execute the following process: acquiring a camera image including at least one analog meter and an object other than the analog meter; acquiring a master image captured at a position closest to the position where the analog meter included in the camera image was captured; identifying a specified analog meter included in the camera image and the master image based on the processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image; and reading the indicated value of the specified analog meter.

[0108] (Appendix 16) A recording medium according to Appendix 15, having recorded thereon a program that causes a computer to execute the following processes: as a correction process according to the processing results of the feature point matching process, a process of adjusting the position of the specified analog meter in the camera image so that it matches the position of the specified analog meter in the master image, thereby obtaining a corrected camera image; a process of cutting out an area including the specified analog meter from the corrected camera image, thereby obtaining a cutout image; and a process of reading the indicated value of the specified analog meter in the cutout image.

[0109] (Appendix 17) A recording medium according to Appendix 16, having recorded thereon a program that causes a computer to further execute, as the correction process, a process of deforming the shape of the specified analog meter in the camera image so that it matches the shape of the specified analog meter in the master image, when the shape of the specified analog meter included in the master image is the same as the shape of the specified analog meter when viewed from the front.

[0110] (Supplementary Note 18) A recording medium according to Supplementary Note 17, having recorded thereon a program that causes a computer to execute a process of reading the indicated value of the specified analog meter in the cutout image by inputting the cutout image into an area extraction model to obtain an extraction result indicating an area in the cutout image that corresponds to the meter needle of the specified analog meter, and performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

[0111] (Appendix 19) A recording medium according to Appendix 16, having recorded thereon a program that causes a computer to further execute, as the correction process, a process of deforming the shape of the specified analog meter in the camera image into a perfect circle when the shape of the specified analog meter included in the master image is a perfect circle.

[0112] (Supplementary Note 20) A recording medium according to Supplementary Note 19, having recorded thereon a program that causes a computer to execute a process of reading the indicated value of the specified analog meter in the cutout image by inputting the cutout image into an area extraction model to obtain an extraction result indicating an area in the cutout image that corresponds to the meter needle of the specified analog meter, and performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

[0113] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments and examples. Various modifications that would be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Each embodiment can be combined with other embodiments as appropriate. Furthermore, some or all of the configurations described in Supplements 2 to 8, which are dependent on Supplement 1, may also be dependent on Supplements 9 and 15 in the same manner as Supplements 2 to 8. Furthermore, without departing from the scope of each of the above-described embodiments, not limited to Supplement 1, Supplement 9, and Supplement 15, some or all of the configurations described as Supplements may also be dependent on various hardware, software, various recording means for recording software, or systems.

[0114] REFERENCE SIGNS LIST 11 Image storage unit 12 Image acquisition unit 13 Correction processing unit 14 Cutout processing unit 15 Area extraction unit 16 Indicated value reading unit 100 Image processing device

Claims

1. An image processing device having: an image acquisition means for acquiring a camera image including at least one analog meter and an object other than the analog meter, and acquiring a master image captured at a position closest to the position at which the analog meter included in the camera image was captured; an identification means for identifying a specified analog meter included in the camera image and the master image based on the processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image; and an indication value reading means for performing processing to read the indication value of the specified analog meter.

2. The image processing device according to claim 1, further comprising: a correction processing means for performing a correction process according to the result of the feature point matching process, by performing a process of adjusting the position of the specified analog meter in the camera image so that it matches the position of the specified analog meter in the master image, thereby obtaining a corrected camera image; and a cutting means for performing a process of cutting out an area including the specified analog meter from the corrected camera image, thereby obtaining a cutout image, wherein the indicated value reading means performs a process to read the indicated value of the specified analog meter in the cutout image.

3. The image processing device according to claim 1, wherein the identification means, when the camera image contains multiple analog meters, identifies the specified analog meter from among the multiple analog meters based on the processing results obtained by performing the feature point matching process.

4. An image processing device as described in claim 2, wherein the correction processing means further performs, as the correction processing, a process of deforming the shape of the specified analog meter in the camera image so that it matches the shape of the specified analog meter in the master image when the shape of the specified analog meter included in the master image is the same as the shape of the specified analog meter when viewed from the front.

5. An image processing device as described in claim 4, further comprising an area extraction means for inputting the cutout image into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, wherein the indicated value reading means reads the indicated value of the specified analog meter in the cutout image by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

6. An image processing device as described in claim 2, wherein the correction processing means further performs processing to deform the shape of the specified analog meter in the camera image into a perfect circle when the shape of the specified analog meter included in the master image is a perfect circle as the correction processing.

7. An image processing device as described in claim 6, further comprising an area extraction means for inputting the cutout image into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, wherein the indicated value reading means reads the indicated value of the specified analog meter in the cutout image by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

8. The image processing device according to claim 1, wherein the camera image and the master image are linked with information relating to the location within a specified facility where the image was acquired.

9. An information processing method executed by a computer, which comprises: acquiring a camera image including at least one analog meter and an object other than the analog meter; acquiring a master image captured at a position closest to the position where the analog meter included in the camera image was captured; performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image, based on the processing results obtained, identifying a specified analog meter included in the camera image and the master image; and performing processing to read the indicated value of the specified analog meter.

10. The image processing method according to claim 9, further comprising the steps of: performing, as a correction process according to the result of the feature point matching process, a process of adjusting the position of the specified analog meter in the camera image so that it matches the position of the specified analog meter in the master image, thereby obtaining a corrected camera image; performing a process of cutting out an area including the specified analog meter from the corrected camera image, thereby obtaining a cutout image; and performing a process to read the indicated value of the specified analog meter in the cutout image.

11. An image processing method as described in claim 10, wherein, when the shape of the specified analog meter included in the master image is the same as the shape of the specified analog meter when viewed from the front, the correction process further includes processing to deform the shape of the specified analog meter in the camera image so that it matches the shape of the specified analog meter in the master image.

12. An image processing method as described in claim 11, wherein the cutout image is input into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, and the indicated value of the specified analog meter in the cutout image is read by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

13. An image processing method as described in claim 10, wherein, when the shape of the specified analog meter included in the master image is a perfect circle, the correction process further includes processing to deform the shape of the specified analog meter in the camera image into a perfect circle.

14. An image processing method as described in claim 13, wherein the cutout image is input into an area extraction model to obtain an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, and the indicated value of the specified analog meter in the cutout image is read by performing processing using coordinates corresponding to the center of rotation of the meter needle in the cutout image, coordinates corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, coordinates corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and coordinates corresponding to the tip of the meter needle obtained based on the extraction result.

15. A recording medium having recorded thereon a program that causes a computer to execute the following process: acquire a camera image that includes at least one analog meter and an object other than the analog meter; acquire a master image that is captured at a position closest to the position where the analog meter included in the camera image was captured; identify a specified analog meter included in the camera image and the master image based on the processing results obtained by performing feature point matching processing on the entire area of ​​the camera image and the entire area of ​​the master image; and read the indicated value of the specified analog meter.

16. A recording medium as described in claim 15, having recorded thereon a program that causes a computer to execute the following processes: as a correction process according to the processing results of the feature point matching process, a process of adjusting the position of the specified analog meter in the camera image so that it matches the position of the specified analog meter in the master image, thereby obtaining a corrected camera image; a process of cutting out an area including the specified analog meter from the corrected camera image, thereby obtaining a cutout image; and a process of reading the indicated value of the specified analog meter in the cutout image.

17. A recording medium as described in claim 16, which stores a program that causes a computer to further execute, as the correction process, a process of deforming the shape of the specified analog meter in the camera image so that it matches the shape of the specified analog meter in the master image, when the shape of the specified analog meter included in the master image is the same as the shape of the specified analog meter when viewed from the front.

18. A recording medium as described in claim 17, having recorded thereon a program that causes a computer to execute a process of reading the indicated value of the specified analog meter in the cutout image by inputting the cutout image into an area extraction model, obtaining an extraction result indicating an area in the cutout image corresponding to the meter needle of the specified analog meter, and performing processing using the coordinate corresponding to the center of rotation of the meter needle in the cutout image, the coordinate corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, the coordinate corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and the coordinate corresponding to the tip of the meter needle obtained based on the extraction result.

19. A recording medium as described in claim 16, which stores a program that causes a computer to further execute, as the correction process, a process of deforming the shape of the specified analog meter in the camera image into a perfect circle when the shape of the specified analog meter included in the master image is a perfect circle.

20. A recording medium as described in claim 19, having recorded thereon a program that causes a computer to execute a process of reading the indicated value of the specified analog meter in the cutout image by inputting the cutout image into an area extraction model, obtaining an extraction result indicating the area in the cutout image corresponding to the meter needle of the specified analog meter, and performing processing using the coordinate corresponding to the center of rotation of the meter needle in the cutout image, the coordinate corresponding to the position of the minimum indicated value of the specified analog meter in the cutout image, the coordinate corresponding to the position of the maximum indicated value of the specified analog meter in the cutout image, and the coordinate corresponding to the tip of the meter needle obtained based on the extraction result.

Images