JP6505034B2 - Character recognition device - Google Patents

Description

  The present invention relates to a technology for recognizing character information of an object such as a power meter.

  An equipment management ledger for electricity meter is created for maintenance inspection of the property. At the time of facility management ledger creation, a character recognition device that recognizes character information indicated by the electric power meter may be used in order to input information of the electric power meter reliably and efficiently.

  Patent Document 1 describes a character recognition device provided with a light-shielding hood, a camera, and an LED light emitter. In the character recognition device described in Patent Document 1, the upper part of the watt-hour meter is covered with a light-shielding hood to block external light, and then the character information indicated by the watt-hour meter is recognized.

Japanese Patent Application Laid-Open No. 10-019518

In the character recognition apparatus described in patent document 1, the irradiation light of LED light-emitting body is reflected by a reflecting plate, and it is directed to a watt-hour meter. Therefore, spots may be formed on the irradiation light, and the accuracy of character recognition may be lowered.
An object of the present invention is to accurately recognize character information of an object such as a power meter.

The character recognition device according to the present invention is
A case with an open front,
An image acquisition apparatus provided on the back of the housing to acquire image data of the front side;
A light source disposed on the side surface of the housing such that an optical axis is directed in a direction inward of the housing and a direction parallel to the front opening surface;
And a character recognition unit that recognizes character information from image information obtained by photographing by the photographing device and stores the character information in a storage device.

  In the aspect of the invention, the light source is disposed on the side surface of the housing such that the light axis is directed in a direction inward of the housing and a direction parallel to the front opening surface. Therefore, light can be applied to the surface of the object without unevenness, and character information of the object can be recognized with high accuracy.

FIG. 1 is a block diagram of a character recognition device 10 according to a first embodiment. FIG. 2 is a cross-sectional view taken along the line A-A ′ of FIG. 1 according to the first embodiment. FIG. 2 is a configuration diagram of a computer 30 according to Embodiment 1. FIG. 2 shows image data according to the first embodiment. 6 is a flowchart showing the operation of the computer 30 according to the first embodiment. FIG. 5 is a block diagram of a character recognition device 10 according to a second embodiment. The figure which shows the image data at the time of light being reflected. FIG. 10 is a configuration diagram of a computer 30 according to a third embodiment. 15 is a flowchart showing an operation of a computer 30 according to a third embodiment. FIG. 10 is a block diagram of a character recognition device 10 according to a fourth embodiment. FIG. 16 is a configuration diagram of a computer 30 according to a fourth embodiment. 15 is a flowchart showing an operation of a computer 30 according to a fourth embodiment. FIG. 16 shows image data according to a fourth embodiment. FIG. 10 is a block diagram of a character recognition device 10 according to a fifth embodiment.

Embodiment 1
In the first embodiment, a character recognition device 10 that recognizes the character information indicated on the name plate 51 of the power meter 50 will be described. However, the character recognition device 10 is not limited to the character information indicated by the watt-hour meter 50, and may be a device that recognizes the character information indicated by the usage meter such as gas or water, etc. It may be a device that recognizes the character information indicated on the object.

*** Description of the configuration ***
The configuration of the character recognition device 10 according to the first embodiment will be described with reference to FIGS. 1 to 3.
As shown in FIG. 1, the character recognition device 10 includes a photographing device 20 and a computer 30.
The imaging device 20 and the computer 30 are connected via a cable and can communicate with each other. The imaging device 20 and the computer 30 may be communicable with each other, and may be connected by another means such as a wireless network.

As shown in FIG. 1, the imaging device 20 includes a housing 21, an image acquisition device 22, a light source 23, and a jig 24.
The housing 21 is open at the front surface 211 side. The image acquisition device 22 is provided on the back surface 212 side of the housing 21 so as to acquire image data of the front surface 211. The image acquisition device 22 is a camera as a specific example. The light source 23 is disposed on the side surface 213 of the housing 21 so that the optical axis is directed in the direction inside the housing 21 and in the direction parallel to the opening surface of the front surface 211. The light source 23 is, as a specific example, an LED or a light bulb. The jig 24 is provided on the side surface 213 on the upper side of the front surface 211 of the housing 21 so as to protrude from the tip on the front surface 211 side of the housing 21.

As shown in FIG. 2, the housing 21 has an inner layer 214 formed of a transmissive diffusion material whose side surface 213 transmits and diffuses light, and an outer layer 215 formed of a blocking material which blocks external light. The transmissive diffusion material is, as a specific example, a white acrylic plate. The blocking material is, as a specific example, a black acrylic plate. The inner layer 214 and the outer layer 215 are fixed to each other by struts 216 formed of an opaque material, which are disposed at the four corners when viewed from the front.
In FIG. 2, an LED which is a light source 23 is provided on a lower side surface 213 of the housing 21. In FIG. 2, six LEDs are arranged upward in a horizontal row on the lower side surface 213. The light emitted from each LED is repeatedly reflected by the inner layer 214 formed of the transmissive diffusion material, and is irradiated to the surface of the watt-hour meter 50.
A spacer 217 is provided on the lower side of the housing 21, and the cable 218 is inserted into the housing 21 from a gap created by the spacer 217 and connected to the LED as the light source 23.
In FIG. 2, the image acquisition device 22 is omitted. Moreover, the dimension of each part of the housing | casing 21 is shown by FIG. With respect to the current watt-hour meter 50, this dimension can improve the accuracy of character recognition, but the present invention is not limited to this dimension, and may be appropriately designed according to the dimension of the object .

As shown in FIG. 3, the computer 30 includes a processor 31, a storage device 32, and an interface 33 as a hardware configuration. The processor 11 is connected to the other hardware such as the storage device 32 and the interface 33 via signal lines, and controls these other hardware.
The processor 31 is an integrated circuit (IC) that performs processing. The processor 11 is a CPU (Central Processing Unit) as a specific example.
The storage device 32 includes a memory 321 and a storage 322. The memory 321 is, as a specific example, a random access memory (RAM). The storage 322 is, as a specific example, a hard disk drive (HDD). The storage 322 may be a portable storage medium such as a Secure Digital (SD) memory card, a CF (Compact Flash), a NAND flash, a flexible disc, an optical disc, a compact disc, a Blu-ray (registered trademark) disc, and a DVD.
The interface 33 is a device for connecting an external device such as the photographing device 20. The interface 33 is a port of Universal Serial Bus (USB) or High-Definition Multimedia Interface (HDMI) as a specific example.

The computer 30 includes a nameplate extraction unit 41, a character recognition unit 42, and an information storage unit 43 as functional components. The character recognition unit 42 includes a meter value acquisition unit 421 and an information acquisition unit 422.
The functions of the nameplate extraction unit 41, the character recognition unit 42, the information storage unit 43, the meter value acquisition unit 421, and the information acquisition unit 422 are realized by software. The storage 322 of the storage device 32 stores programs for realizing the functions of the respective units of the computer 30. The program is read into the memory 321 by the processor 31 and executed by the processor 31. Thereby, the function of each part of the computer 30 is realized.
Note that some or all of the functions of each unit of the computer 30 may be realized by hardware. In this case, the function is a single circuit, a complex circuit, a programmed processor, a parallel programmed processor, a logic IC, a gate array (GA), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA). Is realized by

*** Description of operation ***
The operation of the character recognition device 10 according to the first embodiment will be described with reference to FIG. 1, FIG. 4 and FIG.
The operation of the imaging device 20 according to the first embodiment will be described with reference to FIGS. 1 and 4.
As shown in FIG. 1, in the imaging device 20, the front surface 211 side of the housing 21 is directed to the watt-hour meter 50, and the jig 24 is mounted on the watt-hour meter 50. Then, the imaging device 20 is installed so as to surround the periphery of the watt-hour meter 50 at the front end portion of the housing 21 of the imaging device 20 on the front surface 211 side. In this state, when an image acquisition instruction is input from the user, the light source 23 is turned on, and the image acquisition device 22 acquires image data of the power meter 50. Then, image data as shown in FIG. 4 is acquired.

The operation of the computer 30 according to the first embodiment will be described with reference to FIG.
The process illustrated in FIG. 5 is started when image data is acquired by the imaging device 20.

In the nameplate extraction process of step S11, the nameplate extraction unit 41 extracts nameplate data of an area portion of the nameplate 51 in the image data acquired by the image acquisition device 22.
Specifically, the nameplate extraction unit 41 extracts, as nameplate data, the pixel group of the largest area among the pixel groups in the image data in the luminance range which is predetermined and stored in the storage 322. Alternatively, the nameplate extraction unit 41 performs pattern matching with a reference image of the nameplate 51 which is predetermined and stored in the storage 322, and extracts a pixel group of a region having the highest similarity as nameplate data.

In the counter extraction process of step S12, the meter value acquisition unit 421 of the character recognition unit 42 extracts counter data of the counter area portion in which the meter value is indicated in the nameplate data extracted in step S11.
Specifically, the meter value acquisition unit 421 detects a plurality of areas having many black pixels, and extracts, as counter data, a pixel group of an area having a high likelihood as a counter among the detected areas. The meter value acquisition unit 421 performs the likelihood as a counter for the detected area by at least one of the position of the area on the name plate 51 and the reference image of the counter.

In the numeral detection process of step S13, the meter value acquisition unit 421 detects a numerical value area from the counter data extracted in step S12.
Specifically, the meter value acquisition unit 421 detects the area of the numerical value of the smallest digit by detecting the area of the numeric characters from the area on the right side of the counter data. In some counters, only the area of the lowest order value is different from the area of other order values, and there may be black letters on a white background. Therefore, the meter value acquisition unit 421 performs detection in both cases of white text on a black background and black text on a white background. Then, the meter value acquisition unit 421 detects the character area in order from the area of the numerical value of the smallest digit detected to the left.
The meter value acquisition unit 421 detects a character region using a method such as a projection conversion method that calculates a histogram on the XY coordinate axis for each pixel and detects a rising edge, and a connected component analysis that extracts a connected component of white and black characters. .

In the meter value recognition process of step S14, the meter value acquisition unit 421 acquires a meter value by recognizing characters from the area detected in step S13. The information storage unit 43 writes the acquired meter value in the storage 322.
Specifically, the meter value acquisition unit 421 realizes character recognition by a method such as pattern matching with a numerical image that is predetermined and stored in the storage 322.

In the unit detection process of step S15, the information acquisition unit 422 of the character recognition unit 42 detects a unit of each information in the nameplate data extracted in step S11.
Specifically, the information acquisition unit 422 detects a unit by a method such as pattern matching with a unit image which is predetermined and stored in the storage 322.

In the character detection process of step S16, the information acquisition unit 422 detects a character area around the unit detected in step S15.
Specifically, the information acquisition unit 422 detects a character area in the area around the unit detected in step S15, using the same method as in step S13.

In the character recognition process of step S17, the information acquisition unit 422 acquires character information indicated on the name plate 51 by recognizing characters from the area detected in step S16. The information storage unit 43 writes the acquired character information in the storage 322 in association with the meter value acquired in step S14.
Specifically, the meter value acquisition unit 421 realizes character recognition by a method such as pattern matching with a numerical image and a character image which are predetermined and stored in the storage 322. An image used for pattern matching may be prepared in advance according to the information shown by the power meter 50 and stored in the storage 322.

*** Effect of Embodiment 1 ***
As described above, in the character recognition device 10 according to the first embodiment, the light source 23 is directed to the side surface 213 of the housing 21 in the direction inside the housing 21 and in the direction parallel to the opening surface of the front surface 211 It is arranged so that it faces. Therefore, the light emitted from the light source 23 is repeatedly reflected by the inner layer 214 made of the transmissive diffusion material, and is uniformly irradiated on the surface of the watt-hour meter 50.
As a result, in the image data acquired by the photographing device 20, the light to the name plate 51 portion is not illuminated, or the light source 23 is reflected and it is difficult to cause a situation of so-called overexposure. The accuracy of recognition is enhanced.

Second Embodiment
The second embodiment differs from the first embodiment in that the light source 23 is movable. The second embodiment will explain this difference.

*** Description of the configuration ***
The configuration of the imaging device 20 according to the second embodiment will be described with reference to FIG.
The imaging device 20 includes a moving device 25 in addition to the configuration shown in FIG.
The moving device 25 moves the light source 23 in the front-rear direction of the housing 21 (X direction in FIG. 6). As a specific example, the moving device 25 is a rack and pinion mounted on the light source 23 or a mechanism configured of a ball screw, a rail, and a motor.

*** Description of operation ***
The operation of the imaging device 20 according to the second embodiment will be described with reference to FIGS. 6 and 7.
Similar to the imaging device 20 according to the first embodiment, the image acquisition device 22 acquires image data of the watt-hour meter 50. In this case, depending on the positional relationship between the name plate 51 of the watt-hour meter 50, the image acquisition device 22, and the light source 23, as shown in FIG. 7, the light emitted from the light source 23 may be reflected in the image data. It can not be said that there is no possibility. In FIG. 7, the part where the light was reflected is shown by hatching.
Therefore, when the light emitted from the light source 23 appears in the image data, the moving device 25 operates in accordance with the instruction from the user to move the light source 23. Then, after the light source 23 is moved, image data of the watt-hour meter 50 is acquired by the image acquisition device 22 as in the imaging device 20 according to the first embodiment. Thereby, image data as shown in FIG. 4 is acquired.
The instruction from the user may be input by directly operating a button or the like of the imaging device 20, or may be input via the computer 30.

*** Effect of Embodiment 2 ***
As described above, the character recognition device 10 according to the second embodiment changes the positional relationship between the name plate 51 of the watt-hour meter 50, the image acquisition device 22, and the light source 23 by moving the light source 23. Thereby, light emitted from the light source 23 can be prevented from being reflected in the image data. As a result, the accuracy of character recognition is enhanced.

Third Embodiment
The third embodiment differs from the second embodiment in having a function of controlling the position of the light source 23. In the third embodiment, this difference will be described.

*** Description of the configuration ***
The configuration of the computer 30 according to the third embodiment will be described with reference to FIG.
The computer 30 includes a movement control unit 44 in addition to the configuration shown in FIG. The functions of the movement control unit 44 may be realized by software as well as other functions as well as by hardware.
The movement control unit 34 controls the movement device 25 to set the position of the light source 23 in accordance with how light strikes the name plate 51 of the watt-hour meter 50, which is an object for acquiring image data.

*** Description of operation ***
The operation of the computer 30 according to the third embodiment will be described with reference to FIG.
The process of step S31 is the same as the process of step S11 of FIG. The processes of steps S35 to S40 are the same as the processes of steps S12 to S17 of FIG.

In the image determination process of step S32, the movement control unit 44 determines whether or not the light strikes the nameplate data extracted in step S31 is appropriate.
Specifically, the movement control unit 44 calculates the variance of luminance in the nameplate data, and determines whether the calculated variance is equal to or less than a threshold. The movement control unit 44 determines that the hit of light is appropriate when the dispersion is equal to or less than the threshold, and determines that the hit of light is inappropriate when the dispersion is higher than the threshold.

  In the branching process of step S33, the movement control unit 44 advances the process to step S35 when it is determined in step S32 that the light strikes properly. On the other hand, if it is determined in step S32 that the light strikes are inappropriate, the process proceeds to step S34.

In the light source moving process of step S34, the movement control unit 44 moves the light source 23.
As a specific example, the movement control unit 44 arranges the light source 23 on the back surface 212 side of the housing 21 in the initial state, and moves the light source 23 to the front surface 211 side by the reference distance each time step S34 is executed. Let

*** Effect of Embodiment 3 ***
As described above, in the character recognition device 10 according to the third embodiment, the position of the light source 23 is controlled in accordance with how light strikes the name plate 51. Thereby, it is possible to prevent the light emitted from the light source 23 from being reflected in the image data regardless of the operation of the user. As a result, the accuracy of character recognition is enhanced.

Fourth Embodiment
The fourth embodiment differs from the first embodiment in that image data is acquired while changing the position at which the light source 23 is turned on, and the acquired image data is combined. In the fourth embodiment, this difference will be described.

*** Description of the configuration ***
The configuration of the imaging device 20 according to the fourth embodiment will be described with reference to FIG.
The imaging device 20 differs from the configuration shown in FIG. 1 in that the imaging device 20 includes a plurality of light sources 23 in the front-rear direction of the housing 21.

The configuration of the computer 30 according to the fourth embodiment will be described with reference to FIG.
The calculator 30 includes a lighting control unit 45 and a combining unit 46 in addition to the configuration shown in FIG. The functions of the lighting control unit 45 and the combining unit 46 may be realized by software as well as other functions as well as hardware.
The lighting control unit 45 causes a plurality of image data to be acquired by causing the image acquisition device 22 to acquire image data while changing the position at which the light source 23 is turned on. Here, the lighting control unit 45 changes the position at which the light source 23 is turned on by changing the light source 23 to be turned on.
The combining unit 46 combines a plurality of image data to generate combined data.

*** Description of operation ***
The operation of the computer 30 according to the fourth embodiment will be described with reference to FIGS. 12 and 13.
The process of FIG. 12 is started when an image acquisition instruction is input from the user.

  The processes of steps S44 to S49 are the same as the processes of steps S12 to S17 of FIG.

  In the photographing process of step S41, the lighting control unit 45 causes the image acquisition device 22 to acquire image data while sequentially changing the light source 23 to be lit. Thereby, a plurality of image data acquired in a state where different light sources 23 are lit is acquired.

  In the combining process of step S42, the combining unit 46 sets partial data, which is a part of data of the process target data, according to how the light hits, with each of the plurality of image data acquired in step S41 as the process target data. Extract. Specifically, the combining unit 46 extracts, as partial data, a pixel group in a region where there is no reflection of light emitted from the light source 23 from the luminance for each pixel in the processing target data. Then, the combining unit 46 combines the extracted partial data to generate combined data.

  In the nameplate extraction process of step S43, the nameplate extraction unit 41 extracts the nameplate data of the nameplate 51 portion in the combined data generated in the combining process of step S42.

  It is assumed that the image data U and V shown in FIG. 13 are acquired by the image acquisition device 22. In the image data U, light is reflected around the meter value, and in the image data V, light is reflected around No at the bottom. However, by combining portions of the image data U and V where light is not reflected, it is possible to obtain image data in which light is not reflected in the name plate 51 like the image data W.

*** Effect of Embodiment 4 ***
As described above, the character recognition device 10 according to the fourth embodiment acquires image data while changing the position at which the light source 23 is turned on, and combines the acquired image data. As a result, it is possible to obtain image data in which light strikes the name plate 51 without spots, and the character recognition accuracy is enhanced.

*** Other configuration ***
In the fourth embodiment, the position at which the light source 23 is turned on is changed. However, the number of light sources 23 to be lit may be changed.

  In the fourth embodiment, the imaging device 20 includes the plurality of light sources 23. However, as in the second embodiment, the light source 23 may be moved, and the lighting control unit 45 may move the light source 23 to obtain image data while changing the position at which the light source 23 is turned on.

Embodiment 5
The fifth embodiment differs from the first embodiment in that the computer 30 has an image acquisition device 22. In the fifth embodiment, this difference will be described.

*** Description of the configuration ***
The configuration of the character recognition device 10 according to the fifth embodiment will be described with reference to FIG.
The image acquisition device 22 is not the photographing device 20 but the calculator 30. The computer 30 is a computer in which an image acquisition device 22 such as a smartphone is integrated.
The character recognition device 10 is disposed on the back surface 212 of the housing 21 of the imaging device 20, and the image acquisition device 22 included in the computer 30 is disposed, and the computer 30 is attached to the imaging device 20. A hole is opened in the back surface 212 of the housing 21, and the image acquisition device 22 acquires image data on the front surface 211 side through the hole.

*** Effect of Embodiment 5 ***
As described above, in the character recognition device 10 according to the fifth embodiment, the image acquisition device 22 can be provided on the computer 30 side, and the imaging device 20 and the computer 30 can be integrated. Thereby, the operability of acquisition of image data is enhanced.
Further, when it is not necessary to control the light source 23 via the computer 30, it is not necessary to connect the imaging device 20 and the computer 30 by a cable or the like.

  DESCRIPTION OF SYMBOLS 10 character recognition apparatus, 20 imaging | photography apparatus, 21 housing | casing, 211 front, 212 back surface, 213 side, 22 image acquisition apparatus, 23 light source, 24 jigs, 25 moving apparatuses, 30 calculators, 31 processors, 32 memory | storage devices, 321 memory , 322 storage, 33 interface, 41 name plate extraction unit, 42 character recognition unit, 421 meter value acquisition unit, 422 information acquisition unit, 43 information storage unit, 44 movement control unit, 45 lighting control unit, 46 combining unit, 50 power Meter, 51 name boards.

Claims (8)

A housing having an inner layer formed of a transmissive diffusion material, the side surfaces of which transmit and diffuse light, and an outer layer formed outside the inner layer and formed of a blocking material that blocks external light ; ,
An image acquisition apparatus provided on the back of the housing to acquire image data of the front side;
A light source disposed on the side surface of the housing so that the optical axis is directed in a direction inward of the housing and a direction parallel to the opening surface of the front surface;
A character recognition unit that recognizes character information from image data captured and obtained by the image acquisition device and stores the character information in a storage device. The character recognition device further includes:
The character recognition device according to claim 1 , further comprising a moving device for moving the light source. The character recognition device further includes:
The character recognition device according to claim 2 , further comprising: a movement control unit configured to control the movement device to set the position of the light source in accordance with how light strikes an object for which the image data is to be acquired. The character recognition device further includes:
A lighting control unit that causes a plurality of image data to be acquired by causing the image acquisition device to acquire image data while changing the position at which the light source is turned on;
And a combining unit configured to combine the plurality of image data to generate combined data.
The character recognition device according to any one of claims 1 to 3 , wherein the character recognition unit recognizes character information from composite data generated by the combining unit. The combining unit extracts partial data that is a part of the data to be processed according to how much light is processed, using each of the plurality of image data as processing data, and combines the extracted partial data. The character recognition device according to claim 4 , wherein the synthetic data is generated. The character recognition device according to any one of claims 1 to 5 , wherein a computer having the image acquisition device and the character recognition unit is attached to a photographing device having the housing and the light source. Character recognition device as described. The character recognition device according to any one of claims 1 to 6 , wherein the image acquisition device acquires image data of the object in a state where the periphery of the object is surrounded by the side surface of the housing. The image acquisition device acquires image data of a usage meter in which a meter value representing the usage is indicated;
The character recognition unit according to any one of claims 1 to 7 , wherein the character recognition unit recognizes a meter value in consideration of the fact that the lowest value of the meter values in the image data includes black characters on a white background. Character recognition device as described.

Images