CN108133216B - Nixie tube reading identification method capable of realizing decimal point reading based on machine vision - Google Patents

Nixie tube reading identification method capable of realizing decimal point reading based on machine vision Download PDF

Info

Publication number
CN108133216B
CN108133216B CN201711166125.1A CN201711166125A CN108133216B CN 108133216 B CN108133216 B CN 108133216B CN 201711166125 A CN201711166125 A CN 201711166125A CN 108133216 B CN108133216 B CN 108133216B
Authority
CN
China
Prior art keywords
character
image
nixie tube
reading
characters
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201711166125.1A
Other languages
Chinese (zh)
Other versions
CN108133216A (en
Inventor
陈练
马路
程雷鸣
冯维纲
冯维颖
曹昊
马俊
张国凤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan Zhongyuan Huadian Science & Technology Co ltd
Original Assignee
Wuhan Zhongyuan Huadian Science & Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Wuhan Zhongyuan Huadian Science & Technology Co ltd filed Critical Wuhan Zhongyuan Huadian Science & Technology Co ltd
Priority to CN201711166125.1A priority Critical patent/CN108133216B/en
Publication of CN108133216A publication Critical patent/CN108133216A/en
Application granted granted Critical
Publication of CN108133216B publication Critical patent/CN108133216B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/30—Noise filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/148—Segmentation of character regions
    • G06V30/153—Segmentation of character regions using recognition of characters or words

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Character Input (AREA)

Abstract

The invention relates to a nixie tube reading identification method capable of realizing decimal point reading based on machine vision. The method comprises the following steps: (1) constructing a sample library of single character images of the nixie tube, learning HOG characteristics of the character images in the sample library and generating an SVM classifier; (2) collecting a charactron character image, and extracting a character area in the charactron character image; (3) character segmentation; (4) reading decimal points; (5) the character is identified by a classifier. The invention adopts a classification algorithm combining HOG characteristics and SVM algorithm and a unique decimal reading method, can effectively read the numerical nixie tube reading with decimal points, and has high identification accuracy and good robustness.

Description

Nixie tube reading identification method capable of realizing decimal point reading based on machine vision
Technical Field
The invention belongs to the field of image recognition, and particularly relates to a nixie tube reading recognition method capable of realizing decimal point reading based on machine vision.
Background
The digital instrument is widely applied to various fields of production and life due to the advantages of high precision, easy use and the like. However, reading of digital instruments is mostly achieved manually at present, time is consumed by adopting a manual operation mode, labor intensity is high, efficiency is low, and safety is low in some specific occasions.
The nixie tube identification based on machine vision is to collect the image of the nixie tube by using an image sensor and automatically identify the reading of the nixie tube from the image. At present, a plurality of nixie tube visual identification methods such as a lookup table, a threading method, a neural network and the like exist, but most of the nixie tube visual identification methods aim at the identification of digital characters, and an effective scheme does not exist in a decimal point identification method.
Disclosure of Invention
The invention aims to solve the technical problem of providing a nixie tube reading identification method capable of realizing decimal point reading based on machine vision aiming at the defects of the prior art. The invention adopts a classification algorithm combining HOG characteristics and SVM algorithm and a unique decimal reading method, can effectively read the numerical nixie tube reading with decimal points, and has high identification accuracy and good robustness.
In order to achieve the purpose, the invention adopts the following technical scheme:
the nixie tube reading identification method capable of realizing decimal point reading based on machine vision comprises the following steps:
(1) collecting all single character images of the nixie tube only containing one character, constructing a sample library, reading all the single character sample images in the sample library, extracting the HOG characteristic of each character sample image, and learning and generating an SVM classifier;
(2) collecting a new charactron character image, and extracting a character area in the charactron character image;
(3) performing character segmentation on characters in the extracted charactron character area image to generate a plurality of single character images;
(4) reading decimal points of a character area of the nixie tube;
(5) and respectively identifying a plurality of single character images by using a classifier, and combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube to generate the reading of the nixie tube.
Further, in the nixie tube reading identification method capable of realizing decimal point reading based on machine vision, the nixie tube is a seven-segment nixie tube.
Further, in the nixie tube reading identification method capable of achieving decimal point reading based on machine vision of the present invention, the step of extracting the HOG feature of each character sample image in the step (1) is as follows:
a. preprocessing a character sample image of the nixie tube, namely performing histogram equalization and median filtering;
b. then, carrying out binarization processing on the preprocessed image to obtain a binarized image of the charactron character sample;
c. extracting a character area of the binary image, and generating a character area image which only contains characters and has the same size as the characters;
d. scaling the extracted character area image to a given size, wherein the width and the height of the extracted character area image are 32 pixels by 64 pixels;
e. and extracting HOG characteristics of the zoomed character region image.
Further, in the nixie tube reading identification method capable of realizing decimal point reading based on machine vision of the present invention, the extracting of the character region in the nixie tube character image in the step (2) is to extract a binary image only including the nixie tube character, the gray level value of the character in the binary image is 255, and the gray level value of the background is 0.
Further, in the nixie tube reading identification method capable of achieving decimal point reading based on machine vision of the present invention, the step of single character segmentation in the step (3) is as follows:
a. adjusting the vertical range of the image of the character area of the nixie tube to obtain the coordinates of the upper boundary and the lower boundary of the character area;
b. performing character inclination correction on the inclined character;
c. a single character segmentation is performed for all characters within the region.
Furthermore, in the nixie tube reading identification method capable of achieving decimal point reading based on machine vision according to the present invention, in step a, the vertical range adjustment method includes: projecting the extracted binary image of the charactron character to the vertical direction, wherein the projection formula is as follows:
Figure BDA0001476200270000031
generating a projection histogram by taking the row number of the image as a horizontal axis and the number of pixel points corresponding to the gray value of 255 of each row as a vertical axis, scanning the projection histogram, and extracting the maximum and minimum boundary coordinates of the histogram to further obtain the upper and lower boundaries of the charactron characters;
wherein S isjThe sum of pixel points with the image pixel value of the ith column of 255 is obtained, i and j are respectively row and column coordinates of the pixel, cols is the column width of the image, and P (i, j) takes a value of 0 or 1, when the gray value of the pixel point with the coordinate of (i, j) is 255, P (i, j) takes a value of 1, otherwise, the gray value is 0.
Furthermore, in the nixie tube reading identification method capable of realizing decimal point reading based on machine vision of the present invention, in the step b, the character inclination correction specifically includes: and thinning the extracted binary image of the charactron characters, projecting the thinned image to each direction within +/-60 degrees from the horizontal, selecting the direction of the projection with the largest pixel number in the projection histogram as the direction of the characters, and correcting the characters by using cross-cut transformation.
Furthermore, in the nixie tube reading identification method capable of achieving decimal point reading based on machine vision according to the present invention, in step c, the single character segmentation step is: projecting the corrected character image to the horizontal direction, taking the column number of the image as the horizontal axis, taking the number of pixel points of each corresponding column as the vertical axis to generate a projection histogram, sequentially scanning the projection histogram, sequentially extracting the left and right boundaries of each character, combining the upper and lower boundaries of the character to obtain the region coordinates of each character, and further segmenting the character.
Further, in the nixie tube reading identification method capable of realizing decimal point reading based on machine vision of the present invention, the decimal point reading step in the step (4) is as follows: calculating the maximum character interval of two adjacent characters according to the distance between the two adjacent characters, combining the height of the characters to generate a rectangle with the width being 2/3 times of the maximum character interval and the height being the height of the characters, scanning the whole character area of the nixie tube by using the rectangle according to the sequence from left to right, and when no pixel point exists in the upper 3/4 area and the number of pixel points in the lower 1/4 area is greater than a certain threshold value, determining that the position of the decimal point is located, otherwise, no decimal point exists.
Further, the threshold is calculated according to the edge width of the charactron character.
The threshold calculation method comprises the following steps:
Figure BDA0001476200270000041
where Area is a threshold value, which is the minimum pixel Area of the decimal point region, and L is the pixel width of the character.
Further, in the nixie tube reading identification method capable of reading decimal points based on machine vision of the present invention, in the step (5), the classifier is used to respectively identify a plurality of divided single character images, and the step of generating the reading of the nixie tube by combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube includes the following steps:
a. scaling each segmented character image to a given size with a width and a height of 32 pixels by 64 pixels;
b. extracting HOG characteristics of each zoomed image, and respectively identifying by using an SVM classifier;
c. and then combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube to generate the reading of the nixie tube.
The invention has the beneficial effects that:
the method can effectively identify the seven-segment digital tube reading with decimal point, and has high accuracy and good robustness. The method can replace manual work to realize automatic reading of the nixie tube instrument, greatly improve the working efficiency and reduce the manual cost.
Drawings
FIG. 1 is a flow chart of the present invention.
Detailed Description
To more clearly illustrate the objects, aspects and advantages of the present invention, embodiments of the present invention are described in further detail below with reference to the accompanying drawings.
As shown in fig. 1, a nixie tube reading recognition method based on machine vision and capable of reading decimal points includes steps of utilizing a collected nixie tube character library, utilizing HOG features of character images to learn and generate an SVM classifier for recognizing characters, utilizing sizes and intervals of the characters to generate a rectangular template, and scanning and positioning decimal points, wherein the method includes the following specific implementation steps:
(1) collecting all single character images of the nixie tube only containing one character, constructing a sample library, reading all the single character sample images in the sample library, extracting the HOG characteristic of each character sample image, and learning and generating an SVM classifier;
the method for extracting the HOG characteristics of each character sample image comprises the following steps:
a. preprocessing a character sample image of the nixie tube, namely performing histogram equalization and median filtering;
b. and then carrying out binarization processing on the preprocessed image to obtain a binarization image of the charactron character sample, wherein the binarization method comprises the following steps: graying the image to obtain a grayscale image, and then adopting a large rule method to automatically perform threshold processing.
The formula for graying is:
gray (i, j) ═ 0.299 × R (i, j) +0.587 × G (i, j) +0.144 × B (i, j), i, j are the row and column coordinates of the pixel, respectively; wherein Gray is a Gray value, R, G, B is three color components of red, green and blue respectively;
c. extracting a binary image character area by removing a background and extracting a connected area, and generating a character area image which only contains characters and has the same size as the characters;
d. the extracted character region image is scaled to a given size, and the image is scaled to an image with a width and a height of 32 × 64 in the present embodiment.
e. In this embodiment, the window size of the extracted HOG features is 32 × 64, the block size is 16 × 16, the cell size is 8 × 8, and the number of directions is 9.
(2) Collecting a new charactron character image, and extracting a character area in the charactron character image;
the character region extraction method comprises the following steps: and subtracting the minimum channel value from the maximum channel value of the image to obtain a gray scale image, wherein the digital tube area in the gray scale image is brighter, and the gray scale image is subjected to threshold processing, filtering and area segmentation to obtain a binary image only comprising the digital tube area.
(3) Performing character segmentation on characters in the extracted charactron character area image to generate a plurality of single character images;
the character segmentation method comprises the following steps:
firstly, projecting the extracted binary image of the charactron character to the vertical direction, wherein the projection formula is as follows:
Figure BDA0001476200270000071
generating a projection histogram by taking the row number of the image as a horizontal axis and the number of pixel points corresponding to the gray value of 255 of each row as a vertical axis, scanning the projection histogram, and extracting the maximum and minimum boundary coordinates of the histogram to further obtain the upper and lower boundaries of the charactron characters;
wherein S isjThe gray value of the pixel point with the coordinate (i, j) is 255, i and j are respectively the row and column coordinates of the pixel, cols is the column width of the image, and P (i, j) takes a value of 0 or 1, when the gray value of the pixel point with the coordinate (i, j) is 255, P (i, j) takes a value of 1, otherwise, the gray value is 0.
Secondly, the extracted binary image of the charactron characters is thinned, the thinned image is projected to each direction within the range of +/-60 degrees from the horizontal, the projection direction with the largest pixel number in the projection histogram is selected as the direction of the characters, and then the characters are corrected by the aid of cross-cut transformation.
And finally, projecting the corrected character image in the horizontal direction, taking the column number of the image as a horizontal axis, taking the number of corresponding pixel points in each column as a vertical axis to generate a projection histogram, sequentially scanning the projection histogram, sequentially extracting the left and right boundaries of each character, combining the upper and lower boundaries of the character to obtain the region coordinates of each character, further segmenting the character, and obtaining a plurality of binary images only containing a single character.
(4) Reading decimal points of a character area of the nixie tube;
the decimal point reading method comprises the following steps: calculating the maximum character interval of two adjacent characters according to the distance between the two adjacent characters, combining the height of the characters to generate a rectangle with the width being 2/3 times of the maximum character interval and the height being the height of the characters, scanning the whole character area of the nixie tube by using the rectangle according to the sequence from left to right, and when no pixel point exists in the upper 3/4 area and the number of pixel points in the lower 1/4 area is greater than a certain threshold value, determining that the position of the decimal point is located, otherwise, no decimal point exists.
The threshold is calculated according to the edge width of the charactron characters.
The threshold calculation method comprises the following steps:
Figure BDA0001476200270000081
where Area is the minimum pixel Area (i.e., threshold) of the decimal point region, and L is the pixel width of the character. (5) And respectively identifying a plurality of single character images by using a classifier, and combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube to generate the reading of the nixie tube. The method comprises the following steps:
a. scaling each character image divided to a given size, 64 pixels × 32 pixels;
b. extracting HOG characteristics of each zoomed image, and respectively identifying by using an SVM classifier;
c. and then combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube to generate the reading of the nixie tube.
While the invention has been described in connection with specific embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (9)

1. The nixie tube reading identification method capable of realizing decimal point reading based on machine vision is characterized by comprising the following steps of:
(1) collecting all single character images of the nixie tube only containing one character, constructing a sample library, reading all the single character sample images in the sample library, extracting the HOG characteristic of each character sample image, and learning and generating an SVM classifier;
(2) collecting a new charactron character image, and extracting a character area in the charactron character image;
(3) performing character segmentation on characters in the extracted charactron character area image to generate a plurality of single character images;
(4) reading decimal points of a character area of the nixie tube; the method comprises the following steps: calculating the maximum character interval of two adjacent characters according to the distance between the two adjacent characters, combining the height of the characters to generate a rectangle with the width being 2/3 times of the maximum character interval and the height being the height of the characters, scanning the whole character area of the nixie tube by using the rectangle according to the sequence from left to right, and when no pixel point exists in the upper 3/4 area and the number of pixel points in the lower 1/4 area is greater than a certain threshold value, determining that the position of a decimal point is located, otherwise, no decimal point exists;
(5) and respectively identifying a plurality of single character images by using a classifier, and combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube to generate the reading of the nixie tube.
2. The machine-vision-based nixie tube reading identification method capable of realizing decimal point reading is characterized in that the nixie tube is a seven-segment nixie tube.
3. The nixie tube reading identification method capable of realizing decimal point reading based on machine vision as claimed in claim 1, characterized in that the step of extracting the HOG feature of each character sample image in the step (1) is as follows:
a. preprocessing a character sample image of the nixie tube, namely performing histogram equalization and median filtering;
b. then, carrying out binarization processing on the preprocessed image to obtain a binarized image of the charactron character sample;
c. extracting a character area of the binary image, and generating a character area image which only contains characters and has the same size as the characters;
d. zooming the extracted character area image to a given size;
e. and extracting HOG characteristics of the zoomed character region image.
4. The nixie tube reading identification method capable of reading decimal points based on machine vision as claimed in claim 1, wherein the image after extracting the character area in the nixie tube character image in step (2) is a binary image only containing nixie tube characters, the gray level value of the characters in the binary image is 255, and the gray level value of the background is 0.
5. The nixie tube reading identification method capable of realizing decimal point reading based on machine vision as claimed in claim 1, characterized in that the character segmentation step in the step (3) is as follows:
a. adjusting the vertical range of the image of the character area of the nixie tube to obtain the coordinates of the upper boundary and the lower boundary of the character area;
b. performing character inclination correction on the inclined character;
c. a single character segmentation is performed for all characters within the region.
6. The nixie tube reading identification method capable of realizing decimal point reading based on machine vision according to claim 5, characterized in that in the step a, the vertical range is adjusted specifically as follows: projecting the extracted binary image of the charactron character to the vertical direction, wherein the projection formula is as follows:
Figure DEST_PATH_IMAGE001
generating a projection histogram by taking the row number of the image as a horizontal axis and the number of pixel points corresponding to the gray value of 255 of each row as a vertical axis, scanning the projection histogram, and extracting the maximum and minimum boundary coordinates of the histogram to further obtain the upper and lower boundaries of the charactron characters;
wherein S isjThe sum of pixel points with the image pixel value of 255 in the ith column is represented by i, j is the row and column coordinates of the pixel respectively, cols is the column width of the image, P (i, j) takes a value of 0 or 1, when the gray value of the pixel point with the coordinate of (i, j) is 255,p (i, j) is 1, otherwise it is 0.
7. The nixie tube reading identification method capable of realizing decimal point reading based on machine vision as claimed in claim 5, wherein in the step b, the character tilt correction is specifically: and thinning the extracted binary image of the charactron characters, projecting the thinned image to each direction within +/-60 degrees from the horizontal, selecting the direction of the projection with the largest pixel number in the projection histogram as the direction of the characters, and correcting the characters by using cross-cut transformation.
8. The method for identifying nixie tube reading capable of realizing decimal point reading based on machine vision according to claim 5, wherein in the step c, the single character segmentation step is as follows: projecting the corrected character image to the horizontal direction, taking the column number of the image as the horizontal axis, taking the number of pixel points of each corresponding column as the vertical axis to generate a projection histogram, sequentially scanning the projection histogram, sequentially extracting the left and right boundaries of each character, combining the upper and lower boundaries of the character to obtain the region coordinates of each character, and further segmenting the character.
9. The machine-vision-based nixie tube reading identification method capable of reading decimal points according to claim 1, wherein in the step (5), the classifier is used for respectively identifying the plurality of divided single character images, and the reading of the nixie tube is generated by combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube comprises the following steps:
a. scaling each of the segmented character images to a given size;
b. extracting HOG characteristics of each zoomed image, and respectively identifying by using an SVM classifier;
c. and then combining the pixel coordinate positions of the characters and the decimal points in the character area of the nixie tube to generate the reading of the nixie tube.
CN201711166125.1A 2017-11-21 2017-11-21 Nixie tube reading identification method capable of realizing decimal point reading based on machine vision Active CN108133216B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711166125.1A CN108133216B (en) 2017-11-21 2017-11-21 Nixie tube reading identification method capable of realizing decimal point reading based on machine vision

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711166125.1A CN108133216B (en) 2017-11-21 2017-11-21 Nixie tube reading identification method capable of realizing decimal point reading based on machine vision

Publications (2)

Publication Number Publication Date
CN108133216A CN108133216A (en) 2018-06-08
CN108133216B true CN108133216B (en) 2021-10-12

Family

ID=62388761

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711166125.1A Active CN108133216B (en) 2017-11-21 2017-11-21 Nixie tube reading identification method capable of realizing decimal point reading based on machine vision

Country Status (1)

Country Link
CN (1) CN108133216B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109034160B (en) * 2018-07-06 2019-07-12 江苏迪伦智能科技有限公司 A kind of mixed decimal point digital instrument automatic identifying method based on convolutional neural networks
CN110059693A (en) * 2019-04-18 2019-07-26 华北电力大学（保定） A kind of digital instrument Recognition of Reading system based on wireless sensor network
CN110084241B (en) * 2019-05-05 2023-05-30 山东大学 Automatic ammeter reading method based on image recognition
CN110443220A (en) * 2019-08-13 2019-11-12 树根互联技术有限公司 Digital table image-recognizing method, device, electronic equipment and storage medium
CN110688996A (en) * 2019-09-23 2020-01-14 天津大学 Embedded automatic ruler reading device and method based on visual sensing
CN111222507B (en) * 2020-01-10 2024-03-19 随锐科技集团股份有限公司 Automatic identification method for digital meter reading and computer readable storage medium
CN112348026A (en) * 2020-11-08 2021-02-09 北京工业大学 Magnetic hard disk sequence code identification method based on machine vision
CN113159027B (en) * 2021-04-13 2024-02-09 杭州电子科技大学 Seven-segment digital display instrument identification method based on minimum external rectangular variant

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002158877A (en) * 2000-11-21 2002-05-31 Ricoh Co Ltd Image data processing method and image data processor
CN103984930A (en) * 2014-05-21 2014-08-13 南京航空航天大学 Digital meter recognition system and method based on vision
CN105335745A (en) * 2015-11-27 2016-02-17 小米科技有限责任公司 Recognition method, device and equipment for numbers in images
CN106529537A (en) * 2016-11-22 2017-03-22 亿嘉和科技股份有限公司 Digital meter reading image recognition method
CN106682665A (en) * 2016-12-27 2017-05-17 陕西科技大学 Digital recognition method for seven-segment digital indicator

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002158877A (en) * 2000-11-21 2002-05-31 Ricoh Co Ltd Image data processing method and image data processor
CN103984930A (en) * 2014-05-21 2014-08-13 南京航空航天大学 Digital meter recognition system and method based on vision
CN105335745A (en) * 2015-11-27 2016-02-17 小米科技有限责任公司 Recognition method, device and equipment for numbers in images
CN106529537A (en) * 2016-11-22 2017-03-22 亿嘉和科技股份有限公司 Digital meter reading image recognition method
CN106682665A (en) * 2016-12-27 2017-05-17 陕西科技大学 Digital recognition method for seven-segment digital indicator

Also Published As

Publication number Publication date
CN108133216A (en) 2018-06-08

Similar Documents

Publication Publication Date Title
CN108133216B (en) Nixie tube reading identification method capable of realizing decimal point reading based on machine vision
CN106960208B (en) Method and system for automatically segmenting and identifying instrument liquid crystal number
CN109409366B (en) Distorted image correction method and device based on angular point detection
CN110084241B (en) Automatic ammeter reading method based on image recognition
CN107424142B (en) Weld joint identification method based on image significance detection
CN109426814B (en) Method, system and equipment for positioning and identifying specific plate of invoice picture
CN110210477B (en) Digital instrument reading identification method
CN109409355B (en) Novel transformer nameplate identification method and device
EP2833288A1 (en) Face calibration method and system, and computer storage medium
CN103258198A (en) Extraction method for characters in form document image
CN108920992A (en) A kind of positioning and recognition methods of the medical label bar code based on deep learning
CN102156868A (en) Image binaryzation method and device
CN111353961B (en) Document curved surface correction method and device
CN109993154B (en) Intelligent identification method for single-pointer sulfur hexafluoride instrument of transformer substation
CN111539330B (en) Transformer substation digital display instrument identification method based on double-SVM multi-classifier
CN108985305B (en) Laser etching industrial detonator coded image positioning and correcting method
CN115082466B (en) PCB surface welding spot defect detection method and system
CN104112132A (en) Automatic gun number identification method
CN110674812B (en) Civil license plate positioning and character segmentation method facing complex background
CN112364862B (en) Histogram similarity-based disturbance deformation Chinese character picture matching method
CN111783773A (en) Correction method for angle-oriented inclined wire pole signboard
CN109741273A (en) A kind of mobile phone photograph low-quality images automatically process and methods of marking
CN112348018A (en) Digital display type instrument reading identification method based on inspection robot
CN109753981B (en) Image recognition method and device
CN113642570A (en) Method for recognizing license plate of mine car in dark environment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant