JP2016173760A - Numerical value recognition apparatus, numerical value recognition method, and numerical value recognition program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a numerical value recognition apparatus configured to properly recognize a numerical value indicated by a measurement apparatus, regardless of a difference in imaging field of view, a numerical value recognition method, and a numerical value recognition program.SOLUTION: A numerical value recognition apparatus includes: a storage section which stores patterns of belt-like numeric value strings corresponding to number wheels of a numeric value display section of a measurement apparatus; an extraction section which extracts information indicated by the number wheels for each digit included in an image including numerical value information of the measurement apparatus; a collation section which collates the information indicated by the number wheels with the patterns of the numeric value strings; a recognition section which recognizes a numerical value indicated by the measurement apparatus, on the basis of a collation result of the number wheels collated by the collation section; and a determination section which determines a value of a scale included in the image, on the basis of a collation result of the number wheel located at the lowest digit, collated by the collation section.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a numerical value recognition device, a numerical value recognition method, and a numerical value recognition program for recognizing numerical values indicated by a measuring device such as a watt-hour meter, a gas meter, a distance meter, and a water meter.

  Conventionally, in a measuring device such as a watt-hour meter, a gas meter, a distance meter, and a water meter, a technique for reading a numerical value displayed by a numerical value display unit by image recognition is known (for example, see Patent Document 1). In this technology, image recognition processing is performed using image data captured by the surveillance camera, the numerical part of each digit displayed by the numerical display unit is extracted, and each digit is compared with the numerical part in the model image for verification. After the instruction value is calculated, each digit is corrected to an integer value and the instruction value is read.

Japanese Patent No. 3915973

  In recent years, with the spread of portable terminals such as smartphones, it is conceivable to acquire an instruction value displayed by the numerical display unit by photographing the weighing device with the portable terminal. In this case, there is a possibility that the optical axis at the time of imaging by the user of the portable terminal is inclined without being perpendicular to the surface of the numerical display unit. With respect to this point, the above-described conventional technology is based on the premise that the fixed monitoring camera captures an image of the weighing device, and thus such a difference in imaging field of view has not been taken into consideration.

  The present invention has been made in view of the above, and provides a numerical value recognition device, a numerical value recognition method, and a numerical value recognition program capable of accurately recognizing a numerical value indicated by a weighing device regardless of a difference in imaging field of view. For the purpose.

  In order to solve the above-described problems and achieve the object, the numerical value recognition apparatus according to the present invention includes a plurality of number wheels corresponding respectively to a plurality of digits with a scale consisting of a plurality of scale lines at the lowest position. A numerical value recognition device for recognizing a numerical value indicated by a weighing device based on an image including numerical information of the weighing device that displays a numerical value for measurement by rotating, and is continuous in a strip shape corresponding to the number wheel A storage unit for storing a digit string pattern, an extraction unit for extracting display contents displayed by the digit wheel of each digit included in the image, and a collation for matching the display contents of the digit wheel with the digit string pattern A recognition unit for recognizing an indication value in each digit of the weighing device based on a collation result of the number wheel by the collation unit, and based on a collation result of the lowest digit wheel by the collation unit Included in the image Characterized in that and a discrimination unit for discriminating the value of the scale.

  In the numerical value recognition apparatus according to the present invention, in the above invention, the scale is provided with a long scale line that is displayed longer than the other scale lines at a predetermined interval, and the determination unit is configured to position the long scale line. The scale value is discriminated based on the above.

  In the numerical value recognition apparatus according to the present invention, in the above invention, the determination unit acquires a frequency distribution of pixels corresponding to the scale line in the scale portion, and specifies the position of the long scale line using the frequency distribution. The scale value is determined based on the position of the long scale line and a predetermined reference position.

  The numerical value recognition apparatus according to the present invention is characterized in that in the above-mentioned invention, the numerical value recognition device further includes a calculation unit that performs a calculation on the numerical value recognized by the recognition unit.

  The numerical value recognition apparatus according to the present invention further includes an imaging unit that captures an image of a subject and acquires the image, and a display unit that displays an image acquired by the imaging unit and a recognition result of the recognition unit. It is characterized by having.

  The numerical value recognition apparatus according to the present invention further includes an input unit that receives input of identification information of the weighing device that is a recognition target in the above invention, and the storage unit stores unique information in each of the plurality of weighing devices. The display unit is characterized in that a guide frame indicating a numerical display part of the weighing device corresponding to the identification information received by the input unit is displayed superimposed on the image.

  In the above invention, the numerical value recognition apparatus according to the present invention can communicate with an imaging apparatus that captures an image of a subject and acquires the image via a network, and a communication unit that receives the image transmitted by the imaging apparatus. It is further provided with a feature.

  The numerical value recognition method according to the present invention is a measuring device for displaying a numerical value for measurement by rotating a plurality of number wheels corresponding respectively to a plurality of digits with a scale consisting of a plurality of scale lines at the lowest position. A numerical value recognition method performed by a numerical value recognition device for recognizing a numerical value indicated by the weighing device based on an image including numerical value information, wherein a storage unit stores a pattern of a string of numbers that are continuous in a band shape corresponding to the number wheel The number string pattern is read out from the collation step of collating the display position of the digit wheel of each digit included in the image with the number string pattern, and based on the collation result of the number wheel in the collation step A recognition step for recognizing the indicated value in each digit of the weighing device, and the scale of the scale included in the image based on the collation result of the lowest digit wheel in the collation step. It characterized by having a a determination step of determining the.

  A numerical value recognition program according to the present invention is a measuring device for displaying a numerical value for measurement by rotating a plurality of number wheels respectively corresponding to a plurality of numbers with a scale consisting of a plurality of scale lines at the lowest position. A numerical value recognition device for recognizing a numerical value indicated by the weighing device based on an image including numerical information is stored in a numerical string pattern from a storage unit that stores a pattern of numerical strings connected in a strip shape corresponding to the numerical wheel. A collation step of reading and collating the display position of the digit wheel of each digit included in the image with the pattern of the digit string, and in each digit of the weighing device based on the collation result of the digit wheel in the collation step A recognition step for recognizing an indication value, and a discrimination step for discriminating a value of the scale included in the image based on a collation result of the lowest digit wheel in the collation step. Characterized in that to execute the flop, the.

  According to the present invention, the display content displayed by the digit wheel of each digit of the weighing device included in the image is extracted and collated with a pattern stored in advance, and each of the weighing device is determined based on the collation result of the numeric wheel by the collating unit. In addition to recognizing the indicated value in the digit, the scale unit included in the image is discriminated based on the collation result of the lowest digit wheel by the collation unit. Can be recognized.

FIG. 1 is a block diagram showing a functional configuration of a numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 2 is a diagram illustrating a template (numerical string pattern) stored in the template storage unit of the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 3 is a flowchart showing an outline of processing performed by the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 4 is a diagram showing a display example of the guide frame on the display unit of the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 5 is a diagram showing an outline of numerical value recognition processing performed by the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 6 is a diagram schematically showing a template matching process performed by the matching unit of the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 7 is a flowchart showing an outline of the scale determination process performed by the determination unit of the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 8 is a histogram showing an example of the frequency distribution of black pixels. FIG. 9 is a histogram showing another example of the frequency distribution of black pixels. FIG. 10 is a diagram schematically showing a search range corresponding to the histogram shown in FIG. FIG. 11 is a histogram after the discriminating unit of the numerical value recognition apparatus according to Embodiment 1 of the present invention updates the frequency at the high-frequency position with respect to the histogram shown in FIG. FIG. 12 is a flowchart showing an outline of the numerical value determination process of the scale. FIG. 13 is a diagram (first example) for explaining the outline of the scale instruction value determination process when the two numbers displayed at the lowest place are 9 and 0. FIG. 14 is a diagram (second example) for explaining the outline of the scale instruction value determination process when the two numbers displayed at the lowest place are 9 and 0. FIG. 15 is a diagram showing a display example of recognition results on the display unit of the numerical value recognition apparatus according to Embodiment 1 of the present invention. FIG. 16 is a block diagram illustrating a functional configuration of a system including the numerical value recognition device according to the second embodiment of the present invention. FIG. 17 is a flowchart showing an outline of processing performed by the numerical value recognition apparatus according to Embodiment 2 of the present invention.

  Hereinafter, modes for carrying out the present invention (hereinafter referred to as “embodiments”) will be described.

(Embodiment 1)
FIG. 1 is a block diagram showing a functional configuration of a numerical value recognition apparatus according to Embodiment 1 of the present invention. The numerical value recognition apparatus 1 shown in FIG. 1 is an analog type that displays numerical values for measurement by rotating a plurality of number wheels corresponding to a plurality of digits each having a scale consisting of a plurality of scale lines at the lowest position. This is a device for recognizing a numerical value designated by the weighing device 101 based on an image including numerical information of the weighing device 101. The numerical value recognition apparatus 1 includes an imaging unit 2, an input unit 3, a display unit 4, a storage unit 5, and a control unit 6.

  The imaging unit 2 captures a subject and acquires an image. The imaging unit 2 includes an optical system that collects light from a subject, and a CCD (Charge Coupled Device) or CMOS (Complementary Metal Oxide) that generates and outputs an image signal by photoelectrically converting the light collected by the optical system. Semiconductor) and the like.

  The input unit 3 receives an input of a signal or the like for instructing the operation of the numerical value recognition device 1 and also receives an input of identification information of the weighing device 101. The identification information is, for example, an identification number (ID) written on the surface of the weighing device 101. The input unit 3 is configured using, for example, a touch panel provided so as to overlap the display unit 4.

  The display unit 4 displays various types of information including images captured by the imaging unit 2. The display unit 4 is configured using a display panel such as liquid crystal or organic EL (Electro Luminescence).

The storage unit 5 includes a weighing device information storage unit 51 and a template storage unit 52.
The weighing device information storage unit 51 stores identification information of the weighing device 101 and information related to the meter of the weighing device 101 according to the identification information. The information related to the instrument includes, for example, information related to the numeral wheel of the weighing device 101, information related to the multiplication factor, the presence or absence of a decimal point, and the like.
The template storage unit 52 stores a template that is a numeric string pattern. FIG. 2 is a diagram illustrating a template stored in the template storage unit 52. A template TP shown in the figure shows a pattern of a numeric string in which numeric values of numeric wheels of the weighing device 101 are arranged in a plane.

  The storage unit 5 includes various IC memories such as a ROM (Read Only Memory) or a RAM (Random Access Memory), a hard disk connected with a built-in or data communication terminal, an information recording device such as a CD-ROM, and a reading device thereof. Realized. The storage unit 5 is a program for operating the numerical value recognition device 1 in addition to the image acquired by the imaging unit 2, and a program for the numerical value recognition device 1 to execute various functions, and during execution of this program. Stores data to be used. Specifically, the storage unit 5 stores a numerical value recognition program according to the first embodiment and various parameters such as a threshold value used in numerical value recognition processing.

  Various programs such as a numerical value recognition program recorded by the storage unit 5 can be recorded on a computer-readable recording medium. Recording of various programs in the storage unit 5 or the recording medium may be performed when the computer or the recording medium is shipped as a product, or may be performed by downloading via a communication network. The communication network here is realized by, for example, an existing public line network, a LAN (Local Area Network), a WAN (Wide Area Network) or the like, and may be wired or wireless.

  The control unit 6 controls the overall operation of the numerical value recognition apparatus 1. The control unit 6 includes an image processing unit 61 that performs predetermined image processing such as gain adjustment and A / D conversion on an image signal acquired by the imaging unit 2, and each digit included in the image captured by the imaging unit 2. The extraction unit 62 for extracting the display content displayed by the number wheel, the matching unit 63 for matching the display content of the number wheel with the pattern of the number string, and the weighing device 101 A recognition unit 64 that recognizes an instruction value in each digit, a discrimination unit 65 that discriminates a scale value included in the image based on a collation result of the lowest digit wheel by the collation unit 63, and a numeric value recognized by the recognition unit 64 And an arithmetic unit 66 for performing an operation for multiplying the multiplication factor by 1 and an operation for setting the position of the decimal point.

  The control unit 6 is configured using a CPU (Central Processing Unit) or the like. The control unit 6 reads various programs stored in the storage unit 5 to instruct each unit constituting the numerical value recognition apparatus 1 and transfer data.

  The function of the numerical value recognition apparatus 1 having the above functional configuration is realized using a computer. More specifically, it can be realized by installing predetermined application software on a portable electronic device such as a smartphone, a tablet terminal, or a mobile phone.

  FIG. 3 is a flowchart showing an outline of processing performed by the numerical value recognition apparatus 1. First, the control unit 6 determines whether an input of identification information of the weighing device 101 has been received (step S1). When the input of the identification information of the weighing device 101 is received (step S1: Yes), the numerical value recognition device 1 proceeds to the process of step S2. On the other hand, when the input of identification information is not received (step S1: No), the numerical value recognition apparatus 1 repeats the process of step S1.

  In step S2, the control unit 6 causes the imaging unit 2 to start imaging, and causes the display unit 4 to start displaying an image acquired by the imaging unit 2 through imaging (step S2). At this time, the control unit 6 causes the display unit 4 to display a guide frame corresponding to the size of the numerical display area of the weighing device 101 corresponding to the identification information on the captured image. FIG. 4 is a diagram illustrating a display example of the guide frame on the display unit 4. The display screen 201 shown in FIG. 4 schematically shows a state in which the user of the numerical value recognition apparatus 1 has aligned the numerical display area with the guide frame G. In this way, the display unit 4 displays the guide frame G, so that the user can easily perform alignment with the numerical value display area during imaging.

  The numerical display section shown in FIG. 4 has a format called “one white”. This numerical display section consists of a number wheel with white digits on the black integer part, while black digits are written on the white background only on the lowest digit after the decimal point, and a plurality of scale lines. It consists of a number wheel with a scale. The number wheel of each digit changes the numerical value instructed while rotating upward in the figure according to the amount of measurement object used. Further, in order to read the position of the lowest scale, a reference line 301 is provided on the right side of the scale. As a measuring device, one having a numerical value display part of a format called “all black” without decimal places is also known. In this case, a white number is written on a black background in all the number wheels, and only the lowest number wheel has a scale composed of a plurality of scale lines. The first embodiment can also be applied to a weighing device having such a “full black” numerical value display unit. Needless to say, the present invention is also applicable to a weighing apparatus having a numerical value display unit of another type not described here.

  Thereafter, when the input unit 3 receives an input of a release signal (step S3: Yes), the control unit 6 acquires an image captured by the imaging unit 2 as a still image at a timing when the release signal is received (step S4). ). If the numerical display area is out of the shooting frame during still image shooting, the control unit 6 may display an error display on the display unit 4.

  Subsequently, the numerical value recognition apparatus 1 performs numerical value recognition based on the still image (step S5). FIG. 5 is a diagram showing an outline of the numerical value recognition process performed by the numerical value recognition apparatus 1. First, the extraction unit 62 performs binarization processing for converting the brightness of each pixel into two values of black and white according to a predetermined reference value (step S21). When performing this binarization processing, the extraction unit 62 performs filter processing as necessary.

  Subsequently, the extraction unit 62 extracts a character area to be analyzed (step S22).

  Thereafter, the extraction unit 62 decomposes the character string included in the character area for each character (step S23). For example, the extraction unit 62 detects a valley (portion where no character is described) of the character string by creating a histogram in which the number of black pixels (or white pixels) indicating the character string is integrated along the vertical direction. Disassemble by. Note that the extraction unit 62 may be decomposed by extracting a rectangle circumscribing the character, or may be decomposed by extracting a space between the character strings.

  Subsequently, the extraction unit 62 performs normalization for converting each character into a predetermined size (step S24). At this time, the extraction unit 62 performs a sharpening process, for example. Thereby, the difference in the typeface of the decomposed characters and the difference in shape depending on the shooting angle can be reduced.

  Thereafter, the extraction unit 62 extracts the characteristics of the normalized character (Step S25). Specifically, the extraction unit 62 extracts, for example, features of components in the vertical and horizontal directions and diagonal directions of characters.

  Subsequently, the matching unit 63 performs template matching of each character using the template stored in the template storage unit 52 (step S26). Specifically, the collation unit 63 sequentially searches whether or not the same pattern as the extracted character exists in the template along the search direction (the direction from the top to the bottom in FIG. 2). FIG. 6 is a diagram schematically illustrating template matching processing performed by the matching unit 63. In the case illustrated in FIG. 6, the collation unit 63 collates the character C with the template TP so that the character is a part of 3 (the lower part of the number 3) and a part of 4 (the upper part of the number 4). ). As a result of the collation by the collation unit 63, when the character matches one number, it is determined that the number is described.

  After step S26, the recognition unit 64 performs processing for recognizing integer instruction values for all characters. Hereinafter, processing for one character (steps S27 to S29) will be described.

  First, the recognition unit 64 determines whether or not the character matches one number as a result of template matching (step S27). As a result of template matching, when the character matches one number (step S27: Yes), the recognition unit 64 recognizes the matching number as an instruction value (step S28). As a result of template matching, when the character does not match one number (step S27: No), the recognition unit 64 recognizes the number located above of the two numbers as the instruction value (step S29). For example, in the case illustrated in FIG. 6, the recognition unit 64 recognizes the instruction value of the character C as 3.

  After executing the processing of steps S27 to S29 for all characters, the determination unit 65 determines the value of the scale assigned to the lowest rank (step S30). FIG. 7 is a flowchart showing an outline of the scale determination process performed by the determination unit 65. First, the determination unit 65 creates a histogram that gives a frequency distribution in the left-right direction of black pixels along the vertical direction for the scale portion of the lowest digit wheel (step S41). FIG. 8 is a histogram showing an example of the black pixel frequency distribution in this case. In the histogram shown in FIG. 8, the frequency H at one place is relatively greater than the frequency at another place. A histogram of the type shown in FIG. 8 is obtained when the lowest digit is composed of a number wheel with a scale below the decimal point and a scale is provided, and a long scale line is provided for every 0.1 for the number wheel. FIG. 9 is a histogram showing another example of the frequency distribution of black pixels. In the histogram shown in FIG. 9, the frequencies H1 and H2 of the two places are relatively larger than the frequencies of the other places. A histogram of the type shown in FIG. 9 is obtained when the lowest digit is composed of a number wheel with a scale below the decimal point and a scale is provided, and a long scale line is provided for every 0.05 on the number wheel.

  Subsequently, when there are a plurality of positions (high frequency positions) having a relatively high frequency in the histogram created in step S41 (step S42: Yes), the determination unit 65 determines the frequency search range at the high frequency positions. The frequency is updated to the left by a predetermined number of pixels including a part of the character (number) farthest from the scale area (step S43). FIG. 10 is a diagram schematically showing a search range corresponding to the histogram shown in FIG. FIG. 10 schematically shows a situation in which the search range is expanded to the range E1 at the position corresponding to the high frequency H1, and the search range is expanded to the range E2 at the position corresponding to the high frequency H2. In this case, since the number portion is included in the range E1, the frequency of black pixels increases, while the range E2 is a blank portion between the two upper and lower numbers, and therefore, no black pixel is included. The left end of such ranges E1 and E2 is set, for example, in the approximate center of the lowest digit wheel. This increases the frequency of black pixels for all numbers.

  FIG. 11 is a diagram schematically illustrating a histogram after the determination unit 65 updates the frequency of the high-frequency position in step S43. The frequencies H1 and H2 are updated with respect to the histogram illustrated in FIG. It is a figure which shows a condition typically. In the case shown in FIG. 11, the histogram becomes longer at a place where a character is present (corresponding to 0.0 in FIG. 10, that is, zero position), while black is black at a portion between numbers (corresponding to the position 0.05 in FIG. 10) Since there are no pixels, the length of the histogram does not change. Therefore, the position corresponding to 0.0 and the position corresponding to 0.05 can be clearly distinguished by performing the process of step S43.

  Thereafter, the determination unit 65 sets the position having the highest frequency from the plurality of high-frequency positions updated in step S43 as the zero position (step S44). In the case shown in FIG. 11, the frequency H1 'is maximized, and the position corresponding to this frequency is the zero position.

  In step S42, when a plurality of high-frequency positions do not exist (step S42: No), in other words, when only one high-frequency position exists, the determination unit 65 proceeds to step S44 and sets a zero position.

  After step S44, the determination unit 65 determines a numerical value of the scale (a numerical value one digit lower than the lowest position) according to the positional relationship between the reference line provided on the weighing device 101 and the scale (step S45). . FIG. 12 is a flowchart showing an outline of the numerical value determination process of the scale performed by the determination unit 65. If the lowest character matches one number (step S51: Yes), the determination unit 65 determines that the scale instruction value is 0 (step S52).

  When the lowest character does not match one number (step S51: No), the determination unit 65 determines the vertical positional relationship (vertical relationship) between the reference line and the long scale line indicating the zero position (step). S53). When the long graduation line indicating the zero position is the same as or higher than the position of the reference line (reference position) (step S53: Yes), the determination unit 65 determines the number of the upper number of the two numbers. A difference Δ (us) in the vertical direction between the zero position and the position of the reference line is determined as a scale indication value (step S54). Here, the coordinate difference Δ (us) is calculated in a coordinate system in which the interval between adjacent scale lines is 1. The position of the reference line may be specified based on the captured image, for example. The specifying method at that time may be specified as a predetermined position in the numerical value display unit according to the type of the weighing device 101, or the pixel value of the reference line region portion as in the scale position described above. It may be specified by acquiring the frequency distribution. Furthermore, in the case of a weighing device that is not provided with a reference line, the center position in the vertical direction may be handled as the position of the reference line.

  FIG. 13 is a diagram for explaining the processing in step S54 when the two numbers are 9 and 0. In the case shown in FIG. 13, the long scale line 401 indicating the zero position of the numeral 9 is located above the reference line 301. The determination unit 65 determines the difference Δ (us) = zu−zs between the zero position zu and the reference line position zs in the vertical coordinate z as a scale indication value (corresponding to 0.93 in FIG. 13). Here, the coordinate system z is a coordinate system in which the interval between adjacent scale lines is 1. If Δ (us) has a decimal value, the value of the first decimal place may be rounded or rounded up to an integer value.

  In step S53, when the zero position is located below the position of the reference line (step S53: No), the determination unit 65 determines the zero position of the numeral located below the two numerals and the position of the reference line. A numerical value 10−Δ (sl) obtained by subtracting the difference Δ (sl) from 10 is determined as the scale indication value (step S55). Again, the coordinate difference Δ (sl) is calculated in a coordinate system in which the interval between adjacent graduation lines is 1.

  FIG. 14 is a diagram illustrating the process in step S55 when the two numbers are 9 and 0. In the case shown in FIG. 14, the long scale line 402 indicating the zero position of the numeral 0 is located below the reference line 301. The determination unit 65 determines a numerical value 10−Δ (sl) obtained by subtracting a difference Δ (sl) = zs−zl between the reference line zs and the zero position zl in the vertical coordinate z from 10 as an indication value of the scale ( In FIG. 14, it corresponds to 0.98). When 10−Δ (sl) has a decimal value, the value of the first decimal place may be rounded or rounded up to an integer value.

  After performing the process of step S52, S54 or S55, the determination unit 65 ends the process of step S45, and ends the scale determination process (step S30).

  After step S30 described above, the control unit 6 causes the display unit 4 to display the recognition result by the recognition unit 64 (step S31). FIG. 15 is a diagram illustrating a display example of the recognition result on the display unit 4. The recognition result display screen 211 shown in FIG. 15 displays the recognition result (“617.05” in FIG. 15) converted into numerical data on the result display unit 212 together with the captured still image. Two buttons are displayed below the result display unit 212. The reread button 213 is a button for executing reread when the reading result of the numerical value recognition apparatus 1 is not appropriate. The OK button 214 is a button for the user to approve the reading result. The user determines whether to confirm the reading result by selecting and inputting one of the two buttons.

  When the input unit 3 receives an OK signal input (step S32: Yes), the numerical value recognition apparatus 1 returns to the processing of the main routine shown in FIG. On the other hand, when the input unit 3 does not accept the input of the OK signal and accepts the input of the re-reading instruction signal (step S32: No), the numerical value recognition apparatus 1 returns to the process of step S21. Note that an end button for inputting an instruction signal to end the processing may be further displayed on the recognition result display screen 212. In this case, when the end button is selected by the user, the numerical value recognition device 1 ends the numerical value recognition process.

  Returning to the flowchart shown in FIG. 3, the process performed by the numerical value recognition apparatus 1 after step S5 will be described. The calculation unit 66 performs a calculation based on the acquired numerical value recognition result (step S6). Specifically, the calculation unit 66 calculates the amount of power by multiplying the recognition result by a multiplication factor corresponding to the weighing device 101. For example, in the case of the weighing device 101, as shown in FIG. 4, the multiplication factor (× 600) is described on the display panel. In this case, the calculation unit 66 calculates 61744.05 × 600 to calculate the power amount 3704430 kWh.

  Thereafter, the control unit 6 writes and stores the calculation result calculated by the calculation unit 66 in the storage unit 5 (step S7). Before writing the result calculated by the calculation unit 66 into the storage unit 5, the calculation result may be displayed on the display unit 4 to prompt the user for authentication, and the input unit 3 may accept the input of the authentication result. . Thereby, a series of processes of the numerical value recognition apparatus 1 is completed.

  In step S3, when the input unit 3 does not accept the input of the release signal (step S3: No), the numerical value recognition apparatus 1 proceeds to the process of step S8. In step S8, when the input unit 3 receives an input of an end instruction signal (step S8: Yes), the numerical value recognition apparatus 1 ends a series of processes. On the other hand, when the input unit 3 does not accept the input of the end instruction signal in step S8 (step S8: No), the numerical value recognition apparatus 1 returns to the process of step S3.

  According to the first embodiment of the present invention described above, the display content displayed by the digit wheel of each digit of the weighing device included in the image is extracted and collated with the pattern stored in advance, and the collation of the digit wheel by the collating unit is performed. Based on the result, the indicator value in each digit of the weighing device is recognized, and the scale value included in the image is determined based on the result of matching the lowest digit wheel by the matching unit. It is possible to accurately recognize the numerical value indicated by the weighing device.

  According to the first embodiment, the determination unit determines the scale value based on the position of the long scale line. For this reason, as in the case of the imaging field of view in which the optical axis direction of the optical system included in the imaging unit 2 is not orthogonal to the surface of the weighing device 101, the number of graduations is larger than the assumed number (for example, 10). Even in such a case, it is possible to obtain an accurate scale indication value.

  Further, according to the first embodiment, the image processing apparatus further includes an imaging unit that captures an image of the subject and acquires the image, and a display unit that displays the image acquired by the imaging unit and the recognition result of the recognition unit. For example, it can be realized by a portable terminal. Therefore, an efficient medical examination can be realized by providing the meter reader with the amount of power and gas. In this case, the numerical recognition device 1 may be provided with an output unit such as a printer so that the calculation result is printed and output on paper. By providing the output unit in the numerical value recognition device 1, it becomes possible for a meter reader such as the electric energy and the gas amount to read the electric energy and the gas amount using the numerical value recognition device 1.

  In the first embodiment, the process performed by the recognition unit 64 (steps S27 to S29) and the process performed by the determination unit 65 (step S30) may be executed in reverse order or in parallel. You may make it do.

  In the first embodiment, it is also possible to set a position other than the position of the reference line as the reference position. For example, when the position of the lower end of the area displaying the lowest digit is used as the reference position, the determination unit 65 positions the long scale line at a height that is ½ of the height of the area displaying the lowest digit. In doing so, the indicator value of the scale may be determined as 0.

(Embodiment 2)
FIG. 16 is a block diagram illustrating a functional configuration of a system including the numerical value recognition device according to the second embodiment of the present invention. A numerical value recognition device 11 shown in the figure is connected to the imaging device 12 through a communication network N so that they can communicate with each other. In addition, in FIG. 16, the same code | symbol as FIG. 1 is attached | subjected about the structure part which has the function similar to the numerical value recognition apparatus 1 among the structures with which the numerical value recognition apparatus 11 is provided. Although only one imaging device 12 is illustrated in FIG. 16, it goes without saying that the numerical value recognition device 11 can communicate with a plurality of imaging devices 12 via the communication network N.

  The numerical value recognition device 11 includes a storage unit 5, a control unit 6, and a communication unit 111. The communication unit 111 is a communication interface for transmitting and receiving information to and from the imaging device 12 via the communication network N. The communication network N here is realized by, for example, an existing public line network, LAN, WAN, or the like, regardless of whether it is wired or wireless.

  In the second embodiment, the imaging device 12 has a function of capturing the numerical value of the weighing device 101 and transmitting the captured image and the identification information of the weighing device 101 to the numerical value recognition device 11.

  FIG. 17 is a flowchart showing an outline of processing performed by the numerical value recognition apparatus 11. When the communication unit 111 receives the information of the weighing device 101 transmitted by the imaging device 12 (step S61: Yes), the numerical value recognition device 11 proceeds to the process of step S62. On the other hand, when the communication unit 111 does not receive the information of the weighing device 101 (step S61: No), the numerical value recognition device 11 repeats the process of step S61.

  The processes in steps S62 to S64 sequentially correspond to the processes in steps S5 to S7 shown in FIG. The numerical value recognition process in step S62 is the same as the numerical value recognition process described in Embodiment 1 with reference to FIGS.

  Note that the numerical recognition device 11 may transmit the calculation result to the imaging device 12 after step S64. In this case, the numerical value recognition device 11 may transmit the numerical value information recognized in the numerical value recognition processing in step S62 together.

  According to the second embodiment of the present invention described above, similarly to the first embodiment, the numerical value indicated by the weighing device can be accurately recognized regardless of the difference in the imaging field of view. In addition, an accurate scale indication value can be acquired.

  In addition, according to the second embodiment, numerical recognition is performed based on communication with an imaging device having a communication function, so that calculation can be performed quickly and numerical recognition results obtained from a plurality of imaging devices. Can also be stored together.

  In the second embodiment, the numerical value recognition device 11 may include an input unit and a display unit.

(Other embodiments)
The embodiments for carrying out the present invention have been described so far, but the present invention should not be limited only by the two embodiments described above. For example, the identification information provided on the surface of the weighing device 101 may be imaged by the imaging unit 2, and the identification information may be recognized by recognizing the captured character by image processing.

  Further, as identification information of the weighing device 101, for example, an active element such as a transmitter is provided in the weighing device, and when the numerical value recognition device is set to a predetermined mode and approaches the active element, wireless communication is performed. Thus, identification information may be received and imaging may be started.

  In addition, a recording medium in which identification information is recorded on a barcode or a two-dimensional code is attached to the surface of the weighing device 101, and the identification information is recognized and acquired by the imaging unit 2 photographing the recording medium. May be. In this case as well, the imaging by the imaging unit 2 may be started when the acquisition of the identification information is completed.

  Further, when performing the lowest scale determination process, the storage unit 5 may store a pattern corresponding to the rotation position of the scale in advance, and the determination process may be performed by the matching unit 63 performing pattern matching.

  Further, in the case of a weighing device in which a decimal point is described, such as a numerical value display unit of “one white”, the position of the decimal point may be obtained from an image and determined.

  The calculation may be performed by directly reading the value of the multiplication factor described above the display of the weighing device.

  In the above description, the watt-hour meter has been described as the metering device. However, the metering device is not limited to this, and it goes without saying that the meter can be applied to a gas meter, a distance meter, a water meter, and the like.

  Note that the above-described character and pattern recognition is merely an example, and other known recognition methods can be applied.

  As described above, the present invention can include various embodiments and the like not described herein.

DESCRIPTION OF SYMBOLS 1,11 Numerical value recognition apparatus 2 Imaging part 3 Input part 4 Display part 5 Storage part 6 Control part 12 Imaging device 51 Weighing apparatus information storage part 52 Template storage part 61 Image processing part 62 Extraction part 63 Collation part 64 Recognition part 65 Discrimination part 65 66 Calculation unit 101 Weighing device 111 Communication unit 201 Display screen 211 Recognition result display screen 212 Result display unit 213 Reread button 214 OK button 301 Reference line 401, 402 Long scale line

Claims (9)

Based on an image containing numerical information of a weighing device that displays numerical values for weighing by rotating a plurality of numeral wheels each corresponding to a plurality of digits with a scale consisting of a plurality of scale lines at the lowest level A numerical value recognition device for recognizing a numerical value indicated by the weighing device,
A storage unit that stores a pattern of a string of numbers that are continuous in a band shape corresponding to the number wheel,
An extraction unit for extracting display contents displayed by the number wheel of each digit included in the image;
A collation unit for collating the display content of the numeric wheel with the pattern of the numeric string;
A recognizing unit that recognizes an instruction value in each digit of the weighing device based on a collation result of the number wheel by the collating unit;
A discriminator for discriminating a value of the scale included in the image based on a collation result of the lowest digit wheel by the collator;
A numerical recognition device characterized by comprising: The scale is provided with a long scale line that is displayed longer than the other scale lines at a predetermined interval,
The discrimination unit
The numerical value recognition apparatus according to claim 1, wherein a value of the scale is determined based on a position of the long scale line. The discrimination unit
A frequency distribution of pixels corresponding to the scale line is obtained in the scale portion, the position of the long scale line is specified using the frequency distribution, and the position of the long scale line and a predetermined reference position are used to determine the position of the long scale line. The numerical value recognition apparatus according to claim 2, wherein a scale value is determined.   The numerical value recognition apparatus according to claim 1, further comprising a calculation unit that performs a calculation on the numerical value recognized by the recognition unit. An imaging unit that captures an image of a subject and obtains the image;
A display unit that displays an image acquired by the imaging unit and a recognition result of the recognition unit;
The numerical value recognition apparatus according to claim 1, further comprising: An input unit that receives input of identification information of the weighing device to be recognized;
The storage unit stores information unique to each of the plurality of weighing devices,
6. The numerical value according to claim 5, wherein the display unit displays a guide frame indicating a numerical value display portion of the weighing device corresponding to the identification information received by the input unit in a superimposed manner on the image. Recognition device.   5. A communication unit that is communicable with an imaging device that captures an image of a subject and acquires the image via a network, and further includes a communication unit that receives the image transmitted by the imaging device. The numerical value recognition apparatus as described in any one of. Based on an image containing numerical information of a weighing device that displays numerical values for weighing by rotating a plurality of numeral wheels each corresponding to a plurality of digits with a scale consisting of a plurality of scale lines at the lowest level A numerical value recognition method performed by a numerical value recognition device for recognizing a numerical value indicated by the weighing device,
The number string pattern is read from a storage unit that stores a pattern of number strings connected in a strip shape corresponding to the number wheel, and the display position of the number wheel of each digit included in the image is defined as the number string pattern. A matching step to match;
A recognition step for recognizing an indication value in each digit of the weighing device based on a collation result of the number wheel in the collation step;
A discriminating step for discriminating a value of the scale included in the image based on a collation result of the lowest digit wheel in the collating step;
A numerical value recognition method comprising: Based on an image containing numerical information of a weighing device that displays numerical values for weighing by rotating a plurality of numeral wheels each corresponding to a plurality of digits with a scale consisting of a plurality of scale lines at the lowest level In the numerical value recognition device that recognizes the numerical value indicated by the weighing device,
The number string pattern is read from a storage unit that stores a pattern of number strings connected in a strip shape corresponding to the number wheel, and the display position of the number wheel of each digit included in the image is defined as the number string pattern. A matching step to match;
A recognition step for recognizing an indication value in each digit of the weighing device based on a collation result of the number wheel in the collation step;
A discriminating step for discriminating a value of the scale included in the image based on a collation result of the lowest digit wheel in the collating step;
A numerical recognition program characterized by causing

Images