A coding method for digital character recognition

Abstract

The present invention discloses a coding method for digital character recognition, relates to the field of digital character recognition, and is used to solve the problem that the existing recognition method is difficult to recognize the indication value graphics of low-contrast digital display instruments. The key points of the technical solution are: selecting a template character series, making a normalized template character pattern, using the difference between two template character patterns as a standard template and expressing and encoding it with geometric attributes and difference attributes. The pattern before and after the change of the indication value of the digital display instrument is captured and made into a "difference of the indication value to be recognized" and matched with the standard template, and the coding is used to realize automatic computer matching to obtain the numbers before and after the change, avoiding the binarization of the indication pattern and not being restricted by the contrast of the indication pattern.

Description

Technical Field

[0001] The present invention relates to the field of digital character recognition, and more specifically, to a coding method for digital character recognition. Background Art

[0002] In the prior art, more and more testing / detection / calibration / verification devices and methods obtain the display graphics of the digital instrument being tested, pre-process and binarize the obtained indication value pattern graphics, and then convert them into instrument readings (or indication values) represented by ASCII codes and generate corresponding test reports / detection reports / calibration certificates / verification certificates. Their common feature is that the indication value patterns are processed within the same image plane.

[0003] However, there are low-contrast displays in the displays of the instruments being tested, mainly battery-powered LCD displays. Due to the low contrast of the displays, the cluttered background of the environment, and the imaging of the camera on the surface of the display, the acquired image contains uneven grayscale and has a significant impact on subsequent processing. It is extremely difficult to perform digital character recognition after binarization of such low-contrast indicator value patterns.

[0004] For professional projects in the specific field of testing / detection / calibration / verification, the advantages of taking the model of the instrument being tested as a preset condition and the display of the instrument being tested as prior information are fully utilized, and the area containing the pattern of the digital display instrument indication value can be easily and very simply segmented. The preset conditions and prior information that can be used include the following features;

[0005] (1) When the image acquisition device acquires a set of images containing the indication values ​​of the digital display instrument, the display type of the digital display instrument is a known condition and is associated with the model of the digital display instrument, the geometric dimensions of the display window and the geometric relationship with the digital display instrument are fixed, and the optical characteristics of the display are fixed (for example: ① character high brightness type - LED pen segment / dot matrix / pattern type, LCD display using backlight, etc.); ② character low brightness type - passive reflective LCD pen segment / dot matrix / character type / digital display instrument indication value pattern type, and a display with a low brightness foreground displayed in a high brightness background, etc.); ③ character color difference type - using a color LCD image display, etc.).

[0006] (2) Under stable influence conditions, it is necessary to obtain the indication value of the digital display instrument at intervals. The indication value has a certain regularity and is often evaluated using indicators such as average value and coefficient of variation;

[0007] (3) Change the impact quantity and obtain a new indication value;

[0008] (4) Dose equivalent meters, electricity meters, water meters, natural gas meters and other cumulative quantity measuring instruments (hereinafter collectively referred to as cumulative quantity measuring instruments), whose indicated value is proportional to the time of application of the impact quantity under certain conditions;

[0009] (5) The geometric position of the image acquisition device and the instrument display may remain unchanged during the test / inspection / calibration / verification;

[0010] (6) Parts of the indicator value pattern will appear alternately visible and invisible;

[0011] (7) The testing laboratory is relatively fixed and the same model of the tested instrument will appear multiple times. The laboratory background, the characteristics of the tested instrument and their geometric relationship can be learned from previous images. Not only can the pattern of the digital instrument indication value in the image be cut out, or the image outside the display window be removed, but the character geometric properties of the adaptive indication value characters can also be learned.

[0012] The above features can easily capture the patterns before and after the indication value of the digital display instrument changes and make them into "indication value difference to be identified" and match them with the standard template to obtain the numbers before and after the indication value changes. This method avoids the binarization of the indication value pattern of the digital display instrument.

[0013] Purpose of the present invention:

[0014] Provided is a coding method for digital character recognition, which is used for a computer to automatically match a "difference of an indication value to be recognized" with a standard template, thereby completing the task of digital character recognition in the indication value. Summary of the invention

[0015] The present invention is implemented by the following technical solution, wherein the encoding method for digital character recognition comprises the following steps:

[0016] S1 selects a template character series;

[0017] S2: making a normalized template character pattern, wherein the normalization includes standardizing the size of the template character pattern and unifying the geometric reference points of the template character pattern;

[0018] S3 creates a standard template and represents it with geometric attributes and difference attributes;

[0019] The S4 standard template is represented by coding;

[0020] S5 repeats steps S3 and S4 to enumerate the codes of all standard templates.

[0021] Preferably, in step S1, a template character series is selected, wherein the template character series includes a feature of appearing in the indication value pattern of the same digital display instrument, and the template character series includes any one of S11, S12, and S13:

[0022] S11 seven-segment digital 0-9 character patterns form a template character series, wherein the patterns of characters 0-9 can be equivalently divided into seven segments, and the same segment has the same optical characteristics at the same time, and the display of characters 0-9 is achieved by changing the optical characteristic combination of the segments;

[0023] S12 Arabic numerals 0 to 9 character patterns form a template character series, wherein the patterns of characters 0 to 9 are Arabic numerals;

[0024] S13 The character patterns in the indication value of the same digital display instrument constitute a template character series.

[0025] Preferably, when the normalized template character pattern is prepared in step S2, the normalization includes binarizing the digital display instrument indication value pattern.

[0026] Preferably, in the step S3 series, the template character patterns are represented by character characteristics, and the subtraction of the template character patterns is specifically the difference in character characteristics.

[0027] Preferably, when the standard template is prepared in step S3 and represented by geometric attributes and difference attributes, the difference attributes include at least three categories.

[0028] Further preferably, when S3 prepares a standard template and represents it with geometric attributes and difference attributes, the difference attributes include visible unchanged and hidden unchanged categories.

[0029] Preferably, a duplicate code flag is set when enumerating the codes of all standard templates.

[0030] Preferably, steps S1 to S5 are repeated using at least two template character series.

[0031] Preferably, the codes of all standard templates are enumerated to prepare a standard template code comparison table.

[0032] Further preferably, the indicator value pattern difference to be identified is represented by a code, and the numbers before and after the indicator value changes are queried from a standard template code comparison table.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] In the process of digital character recognition of the indication value pattern captured by the image acquisition device, there is no need to binarize the indication value pattern, get rid of the contrast constraint of the indication value pattern, and use coding to realize the computer's automatic matching of the "indication value pattern difference to be identified" and the standard template. It is also suitable for the recognition of low-contrast and high-contrast digital instrument indication value patterns. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0036] Figure 1 It is a schematic diagram of the process of the present invention.

[0037] Figure 2 It is a schematic diagram of the pattern of the seven-segment digital series 0 to 9 characters.

[0038] Figure 3 It is a schematic diagram of the geometric attribute cutout template of the seven-segment digital series character pattern and a schematic diagram of the position attributes of character characteristics represented by pen segments; Figure 301 is a schematic diagram of the geometric attribute cutout template; Figure 302 is a schematic diagram of the position attributes of character characteristics abstracted as pen segments a, b, c, d, e, f, g; Figure 303 is a schematic diagram of the pen segments abstracted as dots.

[0039] FIG4 is a schematic diagram of subtracting character patterns of a seven-segment digital series template, in which: ■ indicates "a hidden segment changes to a visible segment", and the number 1 in the box indicates that its difference attribute is "1", ◇ indicates "a visible segment changes to a hidden segment", and its difference attribute is "-1" ("-1" is not marked in the figure), □ indicates "no change in visible segment", and the number 0 in the box indicates that its difference attribute is "0", and the "no change in hidden segment" is not shown in the schematic diagram, and its difference attribute is "0"; FIG4(a) is a schematic diagram of the difference attribute after the pattern of the number "0" is subtracted from the patterns of the numbers "1" to "9", FIG4(b) is a schematic diagram of the difference attribute after the pattern of the number "1" is subtracted from the patterns of the numbers "2" to "9", and FIG4(c) is a schematic diagram of the difference attribute after the pattern of the number "2" is subtracted from the patterns of the numbers "3" to "4", respectively. 4(e) is a schematic diagram of the difference attributes after subtracting the pattern of the number “4” from the patterns of numbers “5” to “9” respectively, Figure 4(f) is a schematic diagram of the difference attributes after subtracting the pattern of the number “5” from the patterns of numbers “6” to “9” respectively, Figure 4(g) is a schematic diagram of the difference attributes after subtracting the pattern of the number “6” from the patterns of numbers “7” to “9” respectively, Figure 4(h) is a schematic diagram of the difference attributes after subtracting the pattern of the number “7” from the patterns of numbers “8” to “9” respectively, and Figure 4(i) is a schematic diagram of the difference attributes after subtracting the pattern of the number “8” from the pattern of the number “9”.

[0040] Figure 5 This is a standard template coding comparison diagram showing the conversion of the seven-segment digital character digit A into digit B.

[0041] Figure 6This is a standard template coding comparison diagram showing the conversion of the seven-segment digital character digit B into digit A.

[0042] Figure 7 Schematic diagram of using coding to identify the indication value of a seven-segment digital instrument to be identified, 701 in the figure is a schematic diagram of two indication values ​​of a seven-segment digital instrument to be identified, 702 in the figure is a schematic diagram of the difference of the indication value to be identified, 703 in the figure is a schematic diagram of the coding of the difference of the indication value to be identified, and 704 in the figure is a schematic diagram of querying the result from the standard template coding comparison table.

[0043] Figure 8 This is a schematic diagram of the pattern of the Arabic numerals 0 to 9.

[0044] Fig. 9 The figure is a schematic diagram of the pattern of the digital character "0" indicated by a seven-segment digital display instrument.

[0045] Fig.10 Schematic diagram of the residual image after Arabic numerals 0 to 9 are cut out by "cut out plate 0" to "cut out plate 9" respectively.

[0046] Fig.11 This is a schematic diagram of "the digital character pattern is taken from the indication value of the same digital display instrument" in which the template character pattern adopts the "Arabic numeral series".

[0047] Fig.12 This is a schematic diagram of the effect of filling and bolding the characters 0 to 9 learned from the indication value pattern of the digital display instrument.

[0048] Fig.13 It is a schematic diagram of the positional relationship between the geometric attribute "cutout plate" and the digital display instrument indication value pattern. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0050] [Example 1]

[0051] The present invention is further described in detail below in conjunction with the embodiments. Figure 1 As shown, a coding method for digital character recognition, characterized in that it includes the following steps:

[0052] S1 selects a template character series;

[0053] Generally, character patterns can be divided into many template character series according to factors such as font shape, whether it is bold, whether it is tilted, etc.

[0054] For example: Character patterns can be classified into seven-segment digital series, Arabic numeral series, lower-case Roman numeral series (ⅰ, ⅱ, ⅲ, ⅳ, …, ⅸ), upper-case Roman numeral series (Ⅰ, Ⅱ, Ⅲ, Ⅳ, …, Ⅹ, Ⅺ, Ⅻ), Chinese numerical series (one, two, three, four, five, six, seven, eight, nine, ten), Chinese upper-case numerical series (one, two, three, four, five, six, seven, eight, nine, ten), upper-case English numerical series (ONE, TWO, THREE, …), lower-case English numerical series, and upper- and lower-case English numerical series, etc. Among them, ⅳ, Ⅳ, ONE, etc. are regarded as one character, and so on.

[0055] When performing recognition processing on the pattern of the indication value to be recognized, it is necessary to correspondingly select an appropriate template character series.

[0056] Preferably, in step S1, a template character series is selected. The template character series includes the feature of appearing in the indication value pattern of the same digital display instrument. The template character series includes any one of S11, S12, and S13:

[0057] S11 A template character series is composed of the character patterns of seven-segment digits 0-9. Among them, the patterns of characters 0-9 can be equivalently divided into seven pen segments. The same pen segment has the same optical feature at the same moment, and the display of characters 0-9 is realized by changing the optical feature combination of the pen segments; briefly called the seven-segment digital series;

[0058] S12 A template character series is composed of the character patterns of Arabic numerals 0-9. Among them, the patterns of characters 0-9 are Arabic numerals; briefly called the Arabic numeral series;

[0059] S13 A template character series is composed of the character patterns in the indication value of the same digital display instrument.

[0060] Among them, the most commonly used is the seven-segment digital series. Through abstract expression, the influence of bold / italic can be reduced to the minimum; while the processing methods of the Arabic numeral series and the seven-segment digital series are quite different; the processing method of taking the template character pattern from the indication value of the same digital display instrument may have new differences.

[0061] For example: According to practicality, the (upper / lower-case) Roman numeral series, the (upper / lower-case) Chinese numerical series, and the (upper / lower-case) English numerical series appearing in the "template character series composed of the character patterns in the indication value of the same digital display instrument" have stronger practicality.

[0062] For example: As Fig. 9 shown, the character "0" in the indication value pattern of a certain digital display instrument has a slash in the center which is different from the seven-segment digital series. It is advisable to form a template character series from the character patterns in the indication value of this digital display instrument and define it as the "seven-segment digital series A".

[0063] For another example: the "segment a" of the numerical character "6" in the indication value pattern of a certain digital instrument is visible, or / and the "segment d" of the numerical character "9" is visible, which is different from the seven-segment digital series. It is advisable to form a template character series of the character pattern in the indication value of the digital instrument and revise it to "seven-segment digital series B" or / and "seven-segment digital series C", etc.

[0064] ●The seven-segment digital series includes: traditional pen-segment LED digital tubes for instruments, in which one segment is provided with one LED, and in order to increase the brightness and / or area and / or color characteristics of the LED, multiple LEDs are provided in one segment, and in the dot matrix plane of multiple LEDs, some LEDs are simultaneously "lit" or "extinguished" to be equivalent to being divided into seven segments, and the display of characters 0 to 9 is realized by changing the combination of "lit" or "extinguished" of the seven segments, and the seven-segment characters moving in the dot matrix plane are still classified in the seven-segment digital series.

[0065] ●The seven-segment digital series also includes: traditional instrument pen-segment LCD digital displays, where one segment is provided with an electrode connected to a driving circuit, and the dot-matrix LCD display can simulate the display effect of this pen-segment LCD digital display, and even use color difference to construct a digital character pattern with visible differences between the foreground and background. The pattern can be equivalently divided into seven segments, and each segment has the same optical characteristics at the same time. The combination of the foreground and background of the seven segments forming a visible or blanking state is changed to realize the display of characters 0 to 9. The seven-segment characters moving in the dot-matrix plane are still classified in the seven-segment digital series.

[0066] ●“The same pen segment” includes a set of pixels with the same optical characteristics at the same time, such as: seven-segment digital graphics displayed by LED dot matrix plane, CRT display, dot matrix LCD display, etc.

[0067] S2: making a normalized template character pattern, wherein the normalization includes standardizing the size of the template character pattern and unifying the geometric reference points of the template character pattern;

[0068] The production includes generating a template character pattern using the same font and the same font size, and also includes obtaining a digital display instrument indication value pattern to produce the template character pattern.

[0069] The normalization includes making the template character pattern a standard size and uniformly defining the geometric reference points of the template character pattern when acquiring or constructing the template character pattern. For example, under the same font and the same font size, the "visual" widths of the patterns of the characters 0, 1, and 4 in Arabic numerals are not exactly the same, but when making these patterns, the left and right blanks are adjusted to have the same width. Another example: when making a template character pattern using a digital display instrument indication value pattern, a uniform geometric reference point is defined on the image acquisition device and cut into template character patterns of the same width and height, thereby reducing the amount of calculation for processing the indication value pattern in the future.

[0070] Preferably, when the template character pattern is made using the digital display instrument indication value pattern, the normalization includes binarizing the digital display instrument indication value pattern.

[0071] S3 creates a standard template and represents it with geometric attributes and difference attributes;

[0072] Select two template character patterns from the template character series selected in S1 and subtract them to make a standard template, where the standard template is represented by geometric attributes and difference attributes;

[0073] When processing the pattern recognition of the indication value of a digital instrument, the same geometric properties and difference attribute representation method as the standard template must be selected to process the "indication value pattern difference to be recognized".

[0074] (3.1) Standard templates are represented by geometric attributes

[0075] Example 3-1

[0076] The standard template of the seven-segment digital series is the difference between the character patterns of two templates. The standard template geometric attributes are represented by the distribution positions of the segments "upper, upper right, lower right, lower, lower left, upper left, and middle", which correspond to segments a, b, c, d, e, f, and g respectively.

[0077] Example 3-2

[0078] The standard template of the Arabic numeral series uses 10 bold Arabic numeral characters 0 to 9 as "cutout plates", and the geometric attributes are represented by 10 "cutout plates".

[0079] The representation of the geometrical properties of the standard template also includes reference points defining the standard template and reference points of the template character pattern.

[0080] (3.2) The standard template is represented by the difference attribute;

[0081] Example 3-3

[0082] When 0 is used to represent invisible pixels in a character pattern and 1 is used to represent visible pixels in a character pattern, and when a standard template is made using the difference between two template character patterns, pixel -1 will appear; the difference attribute of the standard template is represented by (0, 1, -1).

[0083] Example 3-4

[0084] Similarly, for black and white character patterns represented by grayscale, usually grayscale value 0 represents black, and grayscale value 255 represents white. When the standard template is made by using the difference between two template character patterns, pixel -255 will appear; the difference attribute of the standard template is represented by (0, 255, -255).

[0085] Example 3-5

[0086] However, it is more practical to use pseudo-color method to make template characters to generate standard templates. For example: template character A is encoded in red RGB (255, 0, 0), template character B is encoded in green RGB (0, 255, 0), the image difference between character A and character B is represented by the sum of RGB, and the common part of character A and character B is yellow RGB (255, 255, 0); the difference attribute of the standard template is represented by {(RGB (255, 0, 0), RGB (0, 255, 0), RGB (255, 255, 0)}. In this way, the character edges can be easily obtained and expanded, and then a "cutout plate" can be made.

[0087] Example 3-6

[0088] When the digital instrument indication value pattern recognition is processed, it is observed from the perspective of the image acquisition device that when the digital instrument indication value pattern change is obtained and the indication value difference to be identified is produced, it can be observed that the indication value difference to be identified has the phenomena of "unchanged", "visible" and "hidden"; therefore, the template character pattern is produced by simulating the above process, and the difference between the two template character patterns is used as the standard template to simulate the indication value difference, and the difference attribute is used to represent the state change of each pen segment, then: the difference attribute of the standard template is represented by "unchanged", "visible" and "hidden".

[0089] Example 3-7

[0090] The Arabic numeral series standard template uses 10 bold Arabic numeral characters 0 to 9 as "cutout plates", which are represented by "cutout plate 0" to "cutout plate 9" respectively. The geometric attributes are represented by 10 "cutout plates"; the residual images after characters 0 to 9 are cut out by "cutout plate 0" to "cutout plate 9" are as follows Fig.10 As shown in the figure, it can be concluded that there is no residual image only when the Arabic numerals are consistent with the "cut-out plate" number. "-1 feature class disappears" or "1 feature class disappears" is the difference attribute of the Arabic numeral series standard template.

[0091] Preferably, when the standard template is prepared in step S3 and represented by geometric attributes and difference attributes, the difference attributes include at least three categories; as shown in Examples 3-3 to 3-6.

[0092] For example, for the difference attribute of the standard template, it is also possible to use the Chinese characters "unchanged", "visible", and "hidden" to represent the three categories:

[0093] ● Use the symbols "0", "+", "-" to indicate the category, or

[0094] ● Use the letters "A", "B", "C" to indicate the category, or

[0095] ● Use integer numbers "0", "1", "-1" to represent the category, or

[0096] ● Use a positive integer number including 0, "1", "2", "0" to represent the category, or

[0097] ● Use 2 binary digits (01)bin, (11)bin, (00)bin to represent the category, or

[0098] ● Use hexadecimal numbers 1Hex, 2Hex, 0Hex to represent the category, or

[0099] ●The difference attribute of the luminous LED uses Chinese characters unchanged, "lit", "off", etc. to indicate the category.

[0100] Further preferably, the difference attributes of the standard template include visible unchanged and hidden unchanged categories.

[0101] When the digital instrument indication value pattern recognition processing is carried out, from the perspective of the image acquisition device, when the digital instrument indication value pattern change is obtained and the indication value difference to be identified is produced, it can also be observed that it includes the "visible unchanged" and "hidden unchanged" categories.

[0102] When the difference between two template character patterns is used as the standard template, the categories of "visible unchanged" and "hidden unchanged" can also be observed.

[0103] For example: the difference attribute of the standard template is represented by the Chinese characters "visible unchanged", "hidden unchanged", "visible", "hidden", or

[0104] ● Use alphabetic symbols to represent differential attributes: "H", "L", "+", "-", or

[0105] ● The difference attribute of the luminous LCD is expressed as "visible unchanged", "blanked unchanged", "lit", "extinguished", or

[0106] ●The differential attribute of the luminous LED is expressed by "unchanged when lit", "unchanged when off", "lit", "off", and so on.

[0107] For example: using the geometric attributes of the standard template of the seven-segment digital series in Example 3-1, and the difference attributes of the standard template in Example 3-3 are represented by (0, 1, -1), then: the standard templates of the seven-segment digital series characters 0 and 3 are represented by geometric attributes and difference attributes (a=0, b=0, c=0, d=0, e=-1, f=-1, g=1).

[0108] Preferably, in the step S3 series, the template character patterns are represented by character characteristics, and the subtraction of the template character patterns is specifically the difference in character characteristics.

[0109] In order to perform logical operations clearly, it is agreed in this article that: from the perspective of the image acquisition device, the image acquisition device records the change of the digital character "A" in the indication value of the digital display instrument to the digital character "B", which is represented by "A jumps to B", and the indication value pattern "B" is subtracted from the indication value pattern "A" to produce the "indication value pattern difference to be identified"; simulating the above process, the template character pattern "B" is subtracted from the template character pattern "A" to produce a standard template when making the code.

[0110] (3.3) Template character pattern represented by character characteristics

[0111] For example, the character pattern of the seven-segment digital series template is represented by character attributes, which include position attributes and segment value attributes;

[0112] ●Location attributes

[0113] According to the relative position of the stroke segments, they are divided into "top, upper right, lower right, bottom, lower left, upper left, and center", and are conventionally composed of (segment a, segment b, segment c, segment d, segment e, segment f, segment g), abbreviated as (a, b, c, d, e, f, g) or (abcdefg); the position attributes of the character features are abstracted as a schematic diagram represented by segments a, b, c, d, e, f, and g, as shown in Figure 302.

[0114] ●Segment value attributes

[0115] There are visible and blanking states. It is agreed that visible is represented by "1" and blanking is represented by "0". The character characteristics of the digital character "0" of the seven-segment digital tube are "a=1, b=1, c=1, d=1, e=1, f=1, g=0";

[0116] The agreed character features are arranged in the order of (abcdefg) and can be abbreviated as (1111110);

[0117] The character characteristics of the numeric character "3" are "a=1, b=1, c=1, d=1, e=0, f=0, g=1", which can be abbreviated as (1111001).

[0118] Character characteristics can also be expressed in hexadecimal (1111110) = 3EHex, (1111001) = 39Hex, or other definitions even in Chinese, such as (1111110) = ("visible", "visible", "visible", "visible", "visible", "visible", "hidden").

[0119] Character attributes may also be specified in the (gfedcba) order or in other order.

[0120] To summarize, character features include both segment position information and segment value information.

[0121] (3.4) Differences in character characteristics

[0122] For example, the difference of the seven-segment digital series character characteristics is the difference of the segment values ​​at the same position. The position attribute of the character characteristics is the same as the geometric attribute of the standard template. The segment is abstracted as a schematic diagram represented by points as shown in Figure 303.

[0123] From the perspective of the image acquisition device, when the image acquisition device records that the indication value character displayed by the seven-segment digital tube jumps from "0" to "3", segments a, b, c, and d remain visible (the segment value is 1), but the segment status changes to "unchanged". It is agreed that "0" is used to represent the "unchanged" state. Coincidentally, the difference attribute between the standard template of character 0 and 3 is also 0;

[0124] Segments e and f change from "visible to invisible", which is represented by "-1" by convention. The difference attribute between the coincidental characters 0 and 3 standard templates is also -1.

[0125] The segment g changes from "hidden to visible", which is represented by "1" by convention. The difference attribute between the coincident character 0 and the standard template 3 is also 1.

[0126] The "to-be-recognized indication value image difference" when the indication value character "0" jumps to "3" can be expressed as (0, 0, 0, 0, -1, -1, 1), which is coincidentally equivalent to "the character characteristic of character 3 minus the character characteristic of character 0", as shown in Table 1:

[0127] Table 1 Comparison table of character characteristics and segment value changes

[0128]

[0129] Similarly, the “difference” of the character “3” changing to “0” can be expressed as (0, 0, 0, 0, 1, 1, -1).

[0130] Rule: In "0 jumps to 3" and "3 jumps to 0", just swap "1" and "-1".

[0131] The difference between two corresponding character attributes is the difference attribute, which is represented by an integer representation (0, 1, -1).

[0132] When the template character pattern is taken from the indication value pattern of a digital display instrument, the segment value of the character characteristic can also be the average grayscale value (or brightness value, or chromaticity value, or R+G+B value, etc.) of the segment. After subtracting the two template character patterns, the difference attributes are classified and processed, and the difference attributes contain at least three categories.

[0133] (3.5) When the character pattern of the "indication value to be recognized" is represented by character characteristics, the methods described in (3.3) and (3.4) can also be used.

[0134] The S4 standard template is represented by coding;

[0135] When performing recognition processing on the indication value pattern to be recognized, the same encoding method must be used.

[0136] Example 4-1: The seven-segment digital series standard template is coded as follows:

[0137] The encoding uses 7 positive integers including zero. One number in the encoding represents one segment, which corresponds to segment a, segment b, segment c, segment d, segment e, segment f, and segment g from left to right. The numbers in the encoding represent difference attributes, and the values ​​of the difference attributes are: 0, 1, 2, where 0 means hidden, 1 means unchanged, and 2 means visible.

[0138] Example 4-2: The Arabic numeral series standard template is coded as follows:

[0139] The code uses ten integers, and the codes correspond to "cutout board 0" to "cutout board 9" from left to right. The coded numbers represent the difference attribute, and the difference attribute values ​​are "-1", "1", and "0", which correspond to "-1 feature class disappears", "1 feature class disappears", and "1 feature class and -1 feature class exist at the same time" respectively.

[0140] When encoding, you can also specify that the codes should be arranged in (gfedcba) order or other order.

[0141] When encoding, the representation of the encoded difference attribute must also be specified;

[0142] For example, when the image acquisition device captures the LED seven-segment digital instrument indication value pattern character "0" jumps to "3", the change process of each pen segment is: a = unchanged, b = unchanged, c = unchanged, d = unchanged, e = off, f = off, g = lit, and the convention is: unchanged = 0, off = -1, lit = 1, then the change process of each pen segment can be expressed as: a = 0, b = 0, c = 0, d = 0, e = -1, f = -1, g = 1;

[0143] Therefore, the above process is simulated to produce the template character patterns of the normalized seven-segment digital characters "0" and "3". The standard template is used to simulate the change process of the indicator value character "0" jumping to "3". The difference attribute is used to represent the state change of each segment. Then:

[0144] (1) The code is expressed in Chinese as “unchanged, unchanged, unchanged, unchanged, off, off, on”; or

[0145] (2) If “unchanged” = 0, “off” = -1, “on” = 1, the code can be expressed as

[0146] "0, 0, 0, 0, -1, -1, 0", or "0 0 0 0 -1 -1 0"; or

[0147] (2) It is specified that the code is represented by a positive integer including 0: no change = 1, "off" = 0, "on" = 2, then the code can be represented as "1, 1, 1, 1, 0, 0, 2", or "1 1 1 1 0 0 2"; or

[0148] (4) If the code is represented by 2 bits of binary, the code can be represented as "01, 01, 01, 01, 00, 00, 10" BIN, or (01010101000010)2; or

[0149] (5) If the hexadecimal code is used instead of the above 2-bit binary code, the code can be expressed as

[0150] "1, 11, 10, 02" Hex, or "1111002" Hex; or

[0151] (6) It is stipulated that: "unchanged" = A, "off" = B, "on" = C, then the code can be expressed as

[0152] "A, A, A, A, B, B, C", or "AAAABBC";

[0153] (7) Provisions: ….

[0154] The standard template is specifically represented by coding

[0155] For example: Using Example 4-1, the seven-segment digital series standard template is represented by code, and the code of characters 0 and 3 is represented as

[0156] (0, 0, 0, 0, -1, -1, 1).

[0157] For example: Using Example 4-2, the Arabic numeral series standard template is represented by code, and the code for characters 0 and 3 is represented as

[0158] (-1, 0, 0, 1, 0, 0, 0, 0, 0).

[0159] When the digital display instrument indication value pattern recognition is processed, from the perspective of the image acquisition device, the code not only includes the geometric characteristics of the indication value pattern difference to be identified on the pure plane map, but also includes the process information of the indication value pattern change.

[0160] S5 repeats steps S3 and S4 to enumerate the codes of all standard templates.

[0161] Preferably, a duplicate code flag is set when enumerating the codes of all standard templates.

[0162] Preferably, at least two template character series are used to repeat steps S1 to S5. For example, firstly, "S11 seven-segment digital 0 to 9 character pattern to form a template character series" is selected to repeat steps S1 to S5, and then "Arabic numerals 0 to 9 character pattern in the indication value of the same digital display instrument to form a template character series" is selected to repeat steps S1 to S5; these two character series form a coding set.

[0163] Preferably, the codes of all standard templates are enumerated to prepare a standard template code comparison table.

[0164] Schematic diagram of the seven-segment digital series standard template coding comparison table, such as Figure 5 and Figure 6 shown.

[0165] Further preferably, the indicator value pattern difference to be identified is represented by a code, and the numbers before and after the indicator value changes are queried from a standard template code comparison table.

[0166] In this document, the number "character 0" distinguishes "uppercase / lowercase letters O / o"; the number "character 1" distinguishes "uppercase / lowercase letters I / i and L / l".

[0167] [Example 2]

[0168] This embodiment is basically the same as Embodiment 1, except that it has the following specific preferences:

[0169] (1) When selecting a template character series, it is preferred to select a seven-segment digital 0-9 character pattern to form a template character series, referred to as a seven-segment digital series. The seven-segment digital series is characterized in that the pattern of characters 0-9 can be equivalently divided into seven segments, and the same segment has the same optical characteristics at the same time. The display of characters 0-9 is achieved by changing the combination of the optical characteristics of the segments. The template character series includes the characteristic of appearing in the indication value pattern of the same digital display instrument.

[0170] (2) producing a normalized template character pattern, wherein the normalization includes standardizing the size of the template character pattern and unifying the geometric reference point of the template character pattern; specifically: producing a standard size, simulating an ideal seven-segment digital tube (primary abstract representation) with horizontal and vertical pen segments as the template character pattern, and realizing the display of characters 0 to 9 by changing the optical feature combination of the pen segments, with the center point of "pen segment g" as the geometric reference point of the template character pattern; preferably, the optical feature of the template character pattern is specifically a black / white pattern, as shown in the schematic diagram Figure 2 As shown; pseudo color representation is more intuitive, but is limited by black and white printing requirements.

[0171] Preferably, the seven-segment digital series characters, the segments can be abstracted to be represented by dots (intermediate abstract representation), such as Figure 3 As shown in 303.

[0172] Further preferably, the pen segment can be abstracted to be represented by letters (a, b, c, d, e, f, g) (high-level abstract representation).

[0173] Common sense refers to the seven-segment digital "segment a" is associated with the "upper horizontal segment", that is, the "geometric attribute".

[0174] (3) Make a standard template and represent it with geometric attributes and difference attributes;

[0175] The standard template of the seven-segment digital series is the difference between the character patterns of two templates. The geometric attributes of the standard template are represented by the distribution positions of the segments "upper, upper right, lower right, lower, lower left, upper left, and middle", which correspond to segments a, b, c, d, e, f, and g respectively, and are abbreviated as a, b, c, d, e, f, and g, respectively. Figure 3 As shown in 302, the stroke segment can be abstracted to be represented by a point, such as Figure 3 As shown in 303.

[0176] Preferably, when a standard template is prepared and represented by geometric attributes and difference attributes, the difference attributes include at least three categories. Specifically, the difference attributes of the standard template include three categories represented by (0, 1, -1) respectively. In the schematic diagram, the difference attribute is represented by ■ to represent "the change of the invisible segment to the visible segment", and the number 1 in the box represents that its difference attribute is "1"; ◇ to represent "the change of the visible segment to the invisible segment", and its difference attribute is "-1" ("-1" is not marked in the figure); □ to represent "the visible segment without change", and the number 0 in the box represents that its difference attribute is "0".

[0177] When the digital display instrument indication value pattern is recognized, the same geometric attribute and difference attribute representation method must be selected to process the "indication value pattern difference to be recognized". In order to extract the geometric attribute from the pattern difference of the seven-segment digital indication value, the following method is used: Figure 3The cutout hole plate shown in 301, wherein the through holes correspond to the pen segments.

[0178] Select two template character patterns from the template character series and then subtract them to make a standard template, and represent the standard template with geometric attributes and difference attributes;

[0179] FIG4 is a schematic diagram of subtracting character patterns of a seven-segment digital series template, in which: ■ indicates "a hidden segment changes to a visible segment", and the number 1 in the box indicates that its difference attribute is "1", ◇ indicates "a visible segment changes to a hidden segment", and its difference attribute is "-1" ("-1" is not marked in the figure), □ indicates "no change in visible segment", and the number 0 in the box indicates that its difference attribute is "0", and the "no change in hidden segment" is not shown in the schematic diagram, and its difference attribute is "0"; FIG4(a) is a schematic diagram of the difference attribute after the pattern of the number "0" is subtracted from the patterns of the numbers "1" to "9", FIG4(b) is a schematic diagram of the difference attribute after the pattern of the number "1" is subtracted from the patterns of the numbers "2" to "9", and FIG4(c) is a schematic diagram of the difference attribute after the pattern of the number "2" is subtracted from the patterns of the numbers "3" to "4", respectively. 4(e) is a schematic diagram of the difference attributes after subtracting the pattern of the number “4” from the patterns of numbers “5” to “9” respectively, Figure 4(f) is a schematic diagram of the difference attributes after subtracting the pattern of the number “5” from the patterns of numbers “6” to “9” respectively, Figure 4(g) is a schematic diagram of the difference attributes after subtracting the pattern of the number “6” from the patterns of numbers “7” to “9” respectively, Figure 4(h) is a schematic diagram of the difference attributes after subtracting the pattern of the number “7” from the patterns of numbers “8” to “9” respectively, and Figure 4(i) is a schematic diagram of the difference attributes after subtracting the pattern of the number “8” from the pattern of the number “9”.

[0180] Preferably, the template character patterns are represented by character characteristics, and the subtraction of the template character patterns is specifically the difference in character characteristics.

[0181] For example, when the segments of the seven-segment digital series are abstracted to be represented by letters (a, b, c, d, e, f, g), the character characteristics of the character "0" are a0=1, b0=1, c0=1, d0=1, e0=1, f0=1, g0=0;

[0182] The character properties of the character "3" are a3=1, b3=1, c3=1, d3=1, e3=0, f3=0, g3=1;

[0183] The subtraction of the character patterns of the character "0" and the character "3" is the difference between the characteristics of the two corresponding characters as follows:

[0184] a 0,3 =a3-a0=0

[0185] b 0,3 =b3-b0=0

[0186] c 0,3 =c3-c0=0

[0187] d 0,3 =d3-d0=0

[0188] e 0,3 =e3-e0=-1

[0189] f 0,3 =f3-f0=-1

[0190] g 0,3 =g3-g0=1

[0191] (4) The standard template is represented by a code;

[0192] Specifically, the code uses 7 positive integers including 0, and the codes correspond to segments a, b, c, d, e, f, and g from left to right; the code value of each segment is: 2, 1, 0, where

[0193] When the segment changes from "hidden to visible", its difference attribute is "1" and the code is "2";

[0194] The state change of the segment is "no change", its difference attribute is "0" and the code is "1";

[0195] The segment changes from "visible to invisible", its difference attribute is "-1" and the encoding is "0";

[0196] (The encoding is equal to the difference attribute plus the number 1, converted into a positive integer representation including 0).

[0197] For example, the encoding of characters 0 and 3 is (1, 1, 1, 1, 0, 0, 2).

[0198] When processing the pattern recognition of the indication value of a digital instrument, the same encoding method must be selected.

[0199] In this embodiment, the encoding equal to the difference attribute is increased by 1 and converted into a positive integer representation including 0.

[0200] (5) Repeat steps 3 and 4 to enumerate the codes of all standard templates.

[0201] Preferably, all the codes of the standard templates are enumerated to make a standard template code comparison table. Figure 5 and Figure 6 As shown, Figure 5This is a standard template coding comparison diagram showing the seven-segment digital character digit A changing to digit B. Figure 6 The standard template coding comparison diagram for seven-segment digital characters, where the digit B is changed to the digit A. The code "2" and the code "0" in Table 5 are interchanged to form Table 6.

[0202] [Example 3]

[0203] This embodiment is basically the same as Embodiment 1, except that it has the following specific preferences:

[0204] (1) Preferably, the template character pattern is taken from the same digital display instrument indication value pattern, such as Fig.11 As shown, the indication value pattern is an Arabic numeral font composed of a dot matrix, that is, the template character pattern adopts the "digital character pattern taken from the same digital display instrument indication value pattern" of the "Arabic numeral series", so as to solve the problem of Arabic numeral font differences.

[0205] (2) Create a standard template and represent it with geometric attributes and difference attributes;

[0206] In this embodiment, the geometric properties of the standard template are represented by ten "cutout plates", which are named "cutout plate 0" to "cutout plate 9" respectively; the specific process of making the "cutout plate" is to fill and bold the characters 0 to 9 learned from the indication value pattern of the digital display instrument, as shown in the schematic diagram Fig.12 Then it is made into a geometric attribute "cut-out plate", and the position relationship diagram of the geometric attribute "cut-out plate" and the digital display instrument indication value pattern is shown as follows Fig.13 shown.

[0207] When the geometric properties of the Arabic numeral series standard template learned from the digital instrument indication value pattern are expressed by "cut-out plate 0" to "cut-out plate 9", there is no residual image only when the Arabic numerals are consistent with the "cut-out plate" number; when the template character pattern difference is established with two Arabic numerals A and B, the standard template is cut out by "cut-out plate A" and the -1 feature class disappears, and the standard template is cut out by "cut-out plate B" and the 1 feature class disappears. "Cut-out plate" is the standard template, and "-1 feature class disappears" or "1 feature class disappears" is the difference attribute of the standard template.

[0208] It can also be equivalently expressed as: the standard template has no -1 feature class after being cut out by "cut out board A", and the standard template has no 1 feature class after being cut out by "cut out board B" (only focusing on the effect after the standard template is cut out by the "cut out board").

[0209] (3) Standard templates are represented by codes, and the codes of all standard templates are enumerated.

[0210] Example SSL3-1

[0211] Preferably, the encoding method of this embodiment is as follows:

[0212] The coding uses two positive numbers including 0, and codes 0 to 9 correspond to "cut-out board 0" to "cut-out board 9" respectively. The codes from left to right correspond to "-1 feature class disappears" and "1 feature class disappears" respectively.

[0213] For example:

[0214] The code for character 0 and character 3 is "03";

[0215] The code for the character 3 and the character 0 is "30".

[0216] Example SSL3-2

[0217] Preferably, the encoding method of this embodiment is as follows:

[0218] The coding is represented by 2 sets, one set is "-1 feature class disappears" and the other set is "1 feature class disappears", and the set elements include the codes corresponding to "cutout board 0" to "cutout board 9" (for example, codes 0 to 9 correspond to "cutout board 0" to "cutout board 9" respectively, codes 10 to 19 correspond to "cutout board 0" to "cutout board 9" respectively, and codes 9 to 0 correspond to "cutout board 0" to "cutout board 9" respectively); a set can also be expressed as "1 group", "1 cluster", etc.

[0219] For example, "-1 feature class disappears" is represented by the letter "A", "1 feature class disappears" is represented by the letter "B", etc., then:

[0220] The encoding of character 0 and character 3 is represented as "A=0, B=3";

[0221] The encoding of character 3 and character 0 is expressed as "A=3, B=0".

[0222] The above is an "explicit" expression of the encoding method.

[0223] When it is changed to the "implicit" expression, "the number of the cutout plate that makes the -1 feature class disappear represents the digital character before the change of the indicator value difference to be identified", and "the number of the cutout plate that makes the 1 feature class disappear represents the digital character after the change of the indicator value difference to be identified". In the "implicit" expression, there are still two elements of "2 sets" and "cutout plate number" to constitute the encoding method. For example, the encoding method can be expressed by the formula for the "-1 feature class disappears" set as follows:

[0224] "Numeric characters before the change of the indication value difference to be identified" = "cut-out board number" + "OFFSET"

[0225] In the formula:

[0226] "OFFSET" = -10, if codes 10 to 19 correspond to "cut-out board 0" to "cut-out board 9" respectively;

[0227] If the code 9 to 0 corresponds to "cut-out board 0" to "cut-out board 9"

[0228] "OFFSET" = 0, if codes 0 to 9 correspond to "cut-out plate 0" to "cut-out plate 9", it means {"digital character before the change of the indicator value difference to be identified" = "cut-out plate number"} is the "implicit" expression of the encoding method; similarly, {"digital character after the change of the indicator value difference to be identified" = "cut-out plate number"} is also the "implicit" expression of the encoding method; it belongs to the linear correspondence under the condition of "encoding using 2 sets" in the encoding method.

[0229] When making a normalized template character pattern, specifically: fix the geometric relationship between the image acquisition device and the digital display instrument, obtain the digital display instrument indication value pattern, define a unified geometric reference point on the image acquisition device and cut it into a template character pattern occupying the same width and height, preferably, binarize the optical characteristics of the indication value pattern, enrich the indication value pattern, and make a normalized template character pattern such as Fig.13 shown.

[0230] [Example 4]

[0231] This embodiment is basically the same as Embodiment 1, except that it has the following specific preferences:

[0232] The coding of all standard templates in Example 1 is enumerated to make a standard template coding comparison table; for example: a schematic diagram of the seven-segment digital series standard template coding comparison table, such as Figure 5 and Figure 6 shown.

[0233] When the image acquisition device captures the pattern before and after the indication value of the seven-segment digital instrument changes and makes it into an "indication value image difference to be identified", the code is obtained from the "indication value image difference to be identified" processing; the "indication value image difference to be identified" is represented by the code, and the numbers before and after the indication value change are queried from the standard template code comparison table.

[0234] Specific example: The schematic diagram of the seven-segment digital instrument indication value to be identified by coding is as follows Figure 7 As shown, 701 is a schematic diagram of two seven-segment digital instrument indication values ​​to be identified, 702 is a schematic diagram of the indication value difference to be identified, 703 is a coding schematic diagram of the indication value difference to be identified, and 704 is a schematic diagram of the result queried from the standard template coding comparison table, and the query result is: the digital instrument indication value changes from 0 to 3.

[0235] [Example 5]

[0236] This embodiment has the following specific preferred options: Step S5 enumerates the codes of all digital standard templates and sets up duplicate code marks.

[0237] Seven-segment digital series Figure 5 and Figure 6 There are duplicate codes in the standard template coding comparison table shown, that is, different digital combinations produce the same standard template, resulting in non-unique results based on the reverse calculation of the code. Figure 5 and Figure 6 The repeated codes in the standard template coding comparison table shown are:

[0238] Character 1 jumps to character 4 or character 7 jumps to character 9 with the same code "1111122".

[0239] Character 4 jumps to character 1 or character 9 jumps to character 7 with the same code "1111100";

[0240] Character 1 jumps to character 7 or character 4 jumps to character 9 with the same code "2111111".

[0241] Character 7 transitions to character 1 or character 9 transitions to character 4 have the same encoding "0111111".

[0242] In the specific field of testing / detection / calibration / verification, for the same digital display instrument being tested, multiple patterns before and after the indication value changes can be easily captured and made into "indication value difference to be identified" to solve this problem.

[0243] For example, the image acquisition device captures the indication value patterns of three different digits (P1, P2, P3) of the seven-segment digital display instrument, and after processing, the following is obtained:

[0244] ⑴The code of P1 changing to P2 is "1111122",

[0245] (2) The code of P2 changing to P3 is "1012211",

[0246] from Figure 5 and Figure 6 The results of the query in the standard template coding comparison table are as follows:

[0247] (1) The change of P1 to P2 may be "character 1 jumps to character 4" or "character 7 jumps to character 9" (P2 may be character "4" or character "9")

[0248] (2) The only option for P2 to change to P3 is "character 4 changes to character 6" (P2 can choose one of the characters "4" and "9", and can only take character "4", excluding the option of "character 7 changes to character 9" when P1 changes to P2);

[0249] The final result is: P3=6, P2=4, P1=1.

[0250] For example, the image acquisition device captures the indication value patterns of three different digits (P1, P2, P3) of the seven-segment digital display instrument. After processing, the code obtained is: the code of P1 changing to P2 is "2111111", and the code of P2 changing to P3 is "1111100".

[0251] from Figure 5 and Figure 6 The results of the query in the standard template coding comparison table are as follows:

[0252] (1) The change of P1 to P2 may be "character 1 jumps to character 7" or "character 4 jumps to character 9" (P2 may be character "7" or "9")

[0253] (2) The change of P2 to P3 can only be "character 4 jumps to character 1" or "character 9 jumps to character 7" (P2 may be character "9" or "4", and combining (1) and (2) it can be concluded that P2 = 9),

[0254] The final result is: P1=4, P2=9, P3=7.

[0255] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A coding method for digital character recognition, characterized in that The following steps are involved: S1 selects a template character series; S2: making a normalized template character pattern, wherein the normalization includes standardizing the size of the template character pattern and unifying the geometric reference points of the template character pattern; S3 creates a standard template and represents it with geometric attributes and difference attributes; The S4 standard template is represented by coding; S5 repeats steps S3 and S4 to enumerate the codes of all standard templates.

2. A coding method for digital character recognition according to claim 1, characterized in that Step S1 selects a template character series, wherein the template character series includes a feature of appearing in the indication value pattern of the same digital display instrument, and the template character series includes any one of S11, S12, and S13: S11 seven-segment digital 0-9 character patterns form a template character series, wherein the patterns of characters 0-9 can be equivalently divided into seven segments, and the same segment has the same optical characteristics at the same time, and the display of characters 0-9 is achieved by changing the optical characteristic combination of the segments; S12 Arabic numerals 0 to 9 character patterns form a template character series, wherein the patterns of characters 0 to 9 are Arabic numerals; S13 The character patterns in the indication value of the same digital display instrument constitute a template character series.

3. A coding method for digital character recognition according to claim 1, characterized in that When the normalized template character pattern is prepared in step S2, the normalization includes binarizing the digital display instrument indication value pattern.

4. A coding method for digital character recognition according to claim 1, characterized in that In the step S3 series, the template character patterns are represented by character characteristics, and the subtraction of the template character patterns is specifically the difference in character characteristics.

5. A coding method for digital character recognition according to claim 1, characterized in that When the standard template is prepared in step S3 and represented by geometric attributes and difference attributes, the difference attributes include at least three categories.

6. A coding method for digital character recognition according to claim 5, characterized in that When S3 creates a standard template and represents it with geometric attributes and difference attributes, the difference attributes include visible unchanged and hidden unchanged categories.

7. A coding method for digital character recognition according to claim 1, characterized in that Set up duplicate code marks when enumerating all the encodings of the standard template.

8. A coding method for digital character recognition according to claim 1, characterized in that Repeat steps S1 to S5 using at least two template character series.

9. A coding method for digital character recognition according to claim 1, characterized in that The codes of all standard templates are enumerated to create a standard template code comparison table.

10. A coding method for digital character recognition according to claim 1 or 9, characterized in that The pattern difference of the indication value to be identified is represented by a code, and the numbers before and after the indication value changes are found out from the standard template code comparison table.

Images