CN113378846A - Bundled bar end face character recognition method - Google Patents
Bundled bar end face character recognition method Download PDFInfo
- Publication number
- CN113378846A CN113378846A CN202110657615.1A CN202110657615A CN113378846A CN 113378846 A CN113378846 A CN 113378846A CN 202110657615 A CN202110657615 A CN 202110657615A CN 113378846 A CN113378846 A CN 113378846A
- Authority
- CN
- China
- Prior art keywords
- face
- character
- image
- area
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Character Input (AREA)
Abstract
The method for recognizing the end face characters of the bundled rods comprises the following steps of: 1, dividing the end face image of the bundled rods into end face images of single rods; step 2, enhancing the character image of the end face of a single bar; step 3, detecting the character area of the end face of a single bar; step 4, dividing the end face characters of the single bar; and 5, identifying the end face characters of the single bar. According to the method, the end face image of the special steel bar bundle is divided into the end face image of the single special steel bar according to the production environment of the special steel bar and the characteristics of the end face code spraying scheme, so that the image enhancement and the character recognition are carried out, and a technical foundation is laid for finally realizing the strategic requirement of the whole process traceability of the single bar.
Description
Technical Field
The invention relates to a character recognition method, in particular to a bundled bar end face character recognition method, and belongs to the field of character recognition.
Background
The special steel bar is an indispensable material in the industries of construction, machinery, automobiles and the like. With the continuous development of modern society and economy, the demand of various industries on steel with different characteristics is increasing day by day, and the annual output of special steel bars is more than 1500 ten thousand tons as the first major country of steel production in China. The special steel bars are various in types and varieties, steel manufacturers need to produce tens of thousands of special steel bars every day, production information marks when the special steel bars are put in storage are in bundles, the special steel bars have important requirements in many fields, and the situation that the conventional bundling mark mode is changed into a single mark mode tends to be great. Therefore, the national department of science and technology has put forward a strategic requirement for realizing the whole process traceability of a single bar in the finishing production of special steel bars.
With the rapid development of machine vision technology, the machine vision technology is applied more and more in the steel production industry. When the steel manufacturing factory finishes the bar, the end face of the bar is marked with codes, and the codes are identified by a machine vision technology, so that specific production information such as production date, furnace batch number, model number and the like of each special steel bar is obtained. Therefore, the method for recognizing the end face characters of the rods in the bundle is provided, and the problem of automatic identity recognition in the whole process traceability process of a single rod is solved.
Disclosure of Invention
Based on the reasons, the invention aims to provide a bundled bar end face character recognition method, which lays a technical foundation for finally realizing the strategy requirement of single bar full-process traceability.
The invention provides a method for recognizing the end face characters of bundled rods according to the characteristics of a rod production environment. The recognition method is characterized in that a bundle of special steel bar images are divided into single special steel bar end face images according to the shape characteristics of the special steel bars; the detection, correction and segmentation of the character area are completed by combining the characteristic that the angle of the character image is random; and creating and training an SVM classifier by using the single character image after segmentation to complete character recognition.
The method for recognizing the end face characters of the bundled rods comprises the following steps:
(1) dividing the end face image of the bundled rods into end face images of single rods;
(2) enhancing the character image of the end face of a single bar;
(3) detecting the character area of the end face of a single bar;
(4) dividing the end face characters of a single bar;
(5) and identifying the end face characters of the single bar.
The invention has the beneficial technical effects that:
according to the characteristics of the special steel bar production environment and the end face code spraying scheme, the end face image of a bundle of special steel bars is divided into single special steel bar end face images through Hough transformation, the special steel bar end face images are enhanced by adopting an image enhancement method based on wavelet transformation, character region screening is completed by combining a MSER method, character region division is completed by adopting edge detection and projection methods, the identification of the end face characters of the special steel bars is completed by adopting an SVM classifier character identification method, and a technical foundation is laid for finally realizing the strategy requirement of single bar full-process traceability.
Drawings
Fig. 1 is a flow chart of the bundled bar end face character recognition method of the invention.
Detailed Description
The specific operation of the present invention will be described with reference to fig. 1.
The method for recognizing the end face characters of the bundled rods comprises the following steps:
(1) dividing the end face image of the bundled rods into end face images of single rods;
(2) enhancing the character image of the end face of a single bar;
(3) detecting the character area of the end face of a single bar;
(4) dividing the end face characters of a single bar;
(5) and identifying the end face characters of the single bar.
The invention discloses a method for identifying end face characters of bundled special steel bars by taking character pattern character identification of end face marks of bundled bars on a special steel bar finishing production line as an example.
The image acquisition system adopted by the invention consists of a large and constant gigabit network camera with the model of MER-500-14GM/C, a lens with the model of M0814-MP2, a blue annular light source with the model of CST-RS12090-B, a light source controller with the model of CST-DPS24120C-4TD and a porphyrizing industrial personal computer (CORE I5-6500(65W) processor, an AIMB-505G2 industrial control mainboard, a 256G 2.5' SATA solid state hard disk and an 8G DDR4 memory).
The specification of the special steel bar processed by the invention is as follows:
the diameter range of a single special steel bar is as follows: 50mm-260 mm;
diameter range of bundled special steel bars: less than 360 mm.
The invention adopts an ink code spraying instrument with a Kandi brand model K68S to spray code, the color of the ink used by the code spraying instrument is white, the character is sprayed and printed on the end surface of the special steel bar, the sprayed and printed character is formed by sequentially arranging square mark points, hundred digits, ten digits, one digits and circular mark points from left to right, the width of the character area is 35mm, and the height of the character is 17 mm.
The invention aims to verify the feasibility of a character recognition method, light conditions of a steel mill are simulated in a laboratory, a plurality of special steel bar bundle end face images are obtained through an image acquisition system, the special steel bar bundle end face images are divided into single pieces in order to recognize character information on the bar end face, and the special steel bar bundle end face character recognition is finally completed by carrying out image enhancement, character area detection, character division, character recognition and other processes on each special steel bar.
1 dividing the end face image of the bundled bar into end face images of single bar
And the method for dividing the end face images of the special steel bars in bundles is to subtract the images after light supplement and the images before light supplement to obtain the end face area of each special steel bar.
When the area selection is carried out on the images after the light supplement, the selection of the end face area of a single special steel bar is completed by adopting Hough transformation circular detection because the end face of the special steel bar is a circular area. Firstly, supplementing light to the end faces of bundles of special steel bars by using a light source, acquiring images by using an image acquisition system, and detecting and dividing the end face area of each special steel bar after the light is supplemented by adopting a Hough transformation circular detection method; and then, subtracting the obtained area of the single special steel bar and the image without light supplement to obtain the end face of the single special steel bar.
And (3) selecting the end face area of the single special steel bar by using an imfindcircle function in an MATLAB environment. The code is as follows:
i ═ immead ('bugueng. bmp'); % reads the end face image of the bundled special steel bar after the light supplement.
T is 0.7; % sets the binarization parameter.
J ═ im2bw (I, T); and (5) binaryzation of the end face image of the% bundled special steel bar.
J ═ bwearopen (J,5000, 8); opening operation is carried out on the image after% binarization.
[ centers, ra, - ] ═ infidcycles (J, [ Rmin Rmax ], 'ObjectPolarity', 'bright', 'Sensitivity',0.965, 'EdgeThreshold', 0.4); % of the total area is subjected to circular detection, Rmin and Rmax are radius ranges of a circular area to be detected, and the total area is obtained by calculating parameters such as the distance between a camera and the end face of the special steel bar, the focal length of the camera and the like.
[ centers, ra ] ═ huofu _ xiuzheng (centers, radii, J); % eliminates the detected circular area of the end face of the special steel bar.
II ═ imread ('weibull.bmp'); % reads the end face image of the bundled special steel bar without light supplement.
[ a, - ]. size (centers); % calculating the number of connected domains.
for i＝1:1:a
rect ═ centers [ i,1] -ra [ i ] centers [ i,2] -ra [ i ] ra [ i ] ]; % sets the rectangular area.
III ═ imcrop [ II, rect ]; % image cropping.
imwrite (III, 'rod' + i + '. bmp'); % save cropped image.
endfor
2 single bar end face character image enhancement
Because the illumination condition of the production environment of the special steel bar is complex and the contrast of the acquired special steel bar end face image is low, the invention adopts an image enhancement method based on wavelet transformation to enhance the special steel bar end face low-contrast image, and the specific steps are as follows:
(1) image wavelet decomposition
The image can be decomposed into low-frequency components and high-frequency components through wavelet transformation, and analysis of the components shows that most of information of the image is reserved in the low-frequency components, and if the high-frequency components exist, the high-frequency components are noise parts of the image. Wavelet decomposition of the image can be accomplished in the MATLAB environment using the dwt2 function, with the code:
[ cA1, cH1, cV1, cD1] ═ dwt2(I, 'harr'); % cA1 is a low-frequency information component after image decomposition, cH1 is a horizontal information component after image decomposition, cV1 is a vertical horizontal information component after image decomposition, and cD1 is a diagonal information component after image decomposition.
(2) Gamma correction is carried out on low-frequency components of the image
The Gamma correction can realize the adjustment of the contrast between the dark area and the bright area of the image by adjusting the parameters. The contrast between the character region and the background region in the low-frequency component is improved by Gamma correction. The code for performing Gamma correction of an image in the MATLAB environment is:
i1 ═ I./255.0; % adjusts the image gray scale value range to 0 to 1.
Igamcocorrection ═ imagjust (I1, [ 01.0 ], [ 01.0 ], 3.5); % Gamma correction is carried out, and a large number of experiments show that the parameter Gamma is the best when the value of the parameter Gamma is 3.5.
Igamcocorrection ═ igamcocorrection · 255; % adjusts the image gray scale value range to 0 to 255.
I3 ═ uint8 (igamcocorrection); the% converted image format facilitates viewing of the display.
(3) Median filtering of image high frequency components
The analysis shows that the noise existing in the high-frequency component of the image is mainly salt-pepper noise, so the invention adopts median filtering to remove the noise. The method is characterized in that the midfilt 2 function is used for finishing median filtering of high-frequency information under the MATLAB environment, and the code is as follows:
i _ med ═ medfilt2(I3, [1,1 ]); the% I3 is the image high frequency component image and [1,1] is the window size.
(4) Image reconstruction and quadratic median filtering
And carrying out image reconstruction on the processed low-frequency component and the processed high-frequency component to obtain a new special steel bar end face image, analyzing the new special steel bar end face image to find that partial noise still exists in the image, and carrying out secondary median filtering to finally complete the enhancement of the special steel bar end face image. Reconstructing each component of the processed image by using an idwt function in an MATLAB environment to obtain a new image, wherein the code is as follows:
ss ═ idwt (cA1, cH1, cV1, cD1, 'harr'); % reconstructs each processed image component.
3 detection of single bar end face character area
Because the end face of the special steel bar is circular, the character area can present any angle in the storage process, the invention adopts the following steps to finish the detection of the end face character area:
(1) initial detection of character regions
The primary detection of the character region is completed by adopting a method based on MSER, and under the MATLAB environment, the code is as follows:
i ═ imread (' zifu. % read character image
mserregines ═ detectmserfeatures (i); % detection of maximally stable extremal region
mserreginsexes ═ vertcat (cell2mat (mserregines. pixellist)); % stores the detected maximum stable extremal region in a pixel list.
mserMask ═ false (size (i)); % defines coefficients as all false MSER regions.
ind ═ sub2ind (size (msermask), mserreginospixels (: 2), mserreginospixels (: 1)); % takes out the coordinate coefficients of the mser region.
msermask (ind) true; % assigns the place of the corresponding coefficient to true, finds the MSER region.
figure；
imshow(mserMask)；
(2) Secondary screening of character regions
The secondary screening of the character area is based on the length-width ratio r of the character areazArea ratio pzCenter coordinate (x) of connected domainz,yz) The screening of the final finished region from the primarily screened region with the characteristics is finally completed, and through a large number of experiments, r is foundz、pzAnd (x)z,yz) The requirements are as follows: r is more than 0.3z＜3、pz＞0.6、|x0-xz|<Δ、|y0-yz|<And delta. Wherein (x)0,y0) And delta is the allowable deviation of the center point of the connected domain and the center point of the frame of the screening area. A large number of experiments show that the delta value can meet the requirement when being between 0 and 5.
4 single bar end face character segmentation
In the character area image obtained by character area detection, the angle of the character is arbitrary, and the character area needs to be corrected before character segmentation, so the invention adopts the following steps to complete the segmentation of the end face character of the bar:
(1) character image correction
According to the special steel bar end surface marking scheme adopted by the invention, the character area image correction is completed according to the square mark points and the circular mark points. Firstly, using an edge function and selecting a Canny operator to complete the detection of edge information in a character region in an MATLAB environment; secondly, filling the detected closed edge, and finishing the selection of the double-mark points according to the area of the connected domain of the double-mark points and the distance of the central point of the connected domain of the double-mark points; and finally, constructing a vector A by taking the square mark point as a starting point and the circular mark point as an end point, constructing a horizontal vector B by passing through the middle point of the connecting line of the two mark points, calculating an included angle theta between the vector A and the vector B from left to right, and rotating the character image by the theta to obtain a corrected image.
A large number of experiments show that the connected domain area S and the distance d between the central points of the double-mark-point connected domain need to satisfy the following conditions when the double-mark-point is selected: d ═ max (d)ij)、20＜S<80, wherein dijAnd i and j are numbers of the two connected domains respectively.
(2) Character segmentation based on projection method
The character segmentation based on projection refers to that the character segmentation is carried out by taking the segmentation points as the boundary, wherein the segmentation points are obtained by analyzing based on the image gray value distribution histogram. Aiming at the corrected character area image, firstly, calculating the horizontal projection of the character area and finishing the horizontal segmentation of the character area; then, calculating the vertical projection of the character area image; finally, according to the width W of the mark pointSign boardRange, Width Range W of character "11The width ranges W of the characters "0", "2", "3", "4", "5", "6", "7", "8", and 9OthersAnd completing vertical segmentation of the character area, and segmenting the character area into single characters. After a large number of experiments, when WSign board、W1、WOthersRespectively satisfy: w is not less than 6Sign board≤10、9≤W1≤11、17≤WOthersWhen the content is less than or equal to 23, the effect is best.
5 Single rod end face character recognition
In order to enable the final character recognition result to have higher accuracy, the invention adopts a character recognition method based on an SVM classifier to complete the character recognition, and the specific steps comprise:
(1) training SVM classifier
In order to improve the creating and training efficiency of the SVM classifier, the character images are normalized in an MATLAB environment, the training of the SVM classifier is completed through a fitceccoc function, a file which is completed by training is saved by using a save function and used for subsequent recognition, and the code is as follows:
clear,clc,close all；
k＝1；
for m＝0:9
for n＝1:150
(ii) f ═ immead (stricat ('nums \ int2str (m),' \ int2str (n), '. bmp')); % read sample image.
(f) guiyihua (f); % images were normalized so that the sample images were of the same size.
xx＝double(f(:))；
P(k,:)＝xx'；
T(k,1)＝m；
k＝k+1；
end
end
net ═ fitceccoc (P, T); % training.
save mynet net; % save trained classifier.
(2) Character recognition using SVM classifier
Normalizing the character image to be recognized, and completing character recognition by using a predict function, wherein the code is as follows:
xs＝[splitfs,points]；
fonts＝'0123456789'；
load mynet; % load trained classifier file.
for m_18＝1:size(xs,2)-1
g＝xs{m_18}；
g＝double(picPretreatment(g))；
lastesult ═ predict (net, g (:'); % character recognition.
endrest ═ string (lastesult); % output recognition results.
end
endresult。
Claims (6)
1. The method for recognizing the end face characters of bundled bars is used for recognizing the marked characters on the end face of each bar in the bundled bars, the color of the characters is white, and the characters are formed by sequentially arranging square mark points, hundred digits, ten digits, unit digits and circular mark points from left to right, and is characterized by comprising the following steps of:
1, dividing the end face image of the bundled rods into end face images of single rods;
step 2, enhancing the character image of the end face of a single bar;
step 3, detecting the character area of the end face of a single bar;
step 4, dividing the end face characters of the single bar;
and 5, identifying the end face characters of the single bar.
2. The method for recognizing the end face characters of the bundled rods according to claim 1, wherein the method for segmenting the end face images of the bundled rods in the step 1 is to subtract the images after light supplement and the images before light supplement to obtain the end face area of each special steel rod, and the selection of the end face area of a single rod is completed by Hough transform circular detection, and the specific steps are as follows: (1) supplementing light to the end faces of the bundled special steel bars by using a light source, acquiring images by using an image acquisition system, and detecting and segmenting the end face area of each special steel bar after the light is supplemented by adopting a Hough transformation circular detection method; (2) and carrying out subtraction operation on the obtained single special steel bar area and the non-light-supplemented image to obtain the end face of the single special steel bar, and finishing the selection of the end face area of the single special steel bar by using an imfindcircle function in an MATLAB environment.
3. The method for recognizing the end face characters of the rods in the bundle according to claim 1, wherein the step 2 of image enhancement comprises the following specific steps:
(1) decomposing a single bar image into a low-frequency component and a high-frequency component through wavelet transformation;
(2) carrying out Gamma correction on the low-frequency components of the single bar image;
(3) carrying out median filtering on the high-frequency components of the single bar image;
(4) and (5) reconstructing an image of a single bar.
4. The method for recognizing the end face characters of the rods in the bundle according to claim 1, wherein the method for detecting the end face character areas in the step 3 comprises the following specific steps:
(1) completing primary detection of a character area by adopting a MSER-based method;
(2) secondary screening of character area according to length-width ratio r of character areazArea ratio pzAnd connected domain center coordinates (x)z,yz) Final region screening from the initially screened region of features, rz、pzAnd (x)z,yz) The requirements are as follows: r is more than 0.3z＜3、pz＞0.6、|x0-xz|<Δ、|y0-yz|<Δ, wherein (x)0,y0) Is the central coordinate of the frame of the screening area, delta is the allowable deviation of the center point of the connected area and the center point of the frame of the screening area, and the value range of delta is [0,5 ]]。
5. The method for recognizing the end face characters of the rods in the bundle according to claim 1, wherein the method for segmenting the end face characters of the single rod in the step 4 comprises the following specific steps:
(1) correcting character images, specifically comprising the following steps: firstly, using an edge function and selecting a Canny operator to complete the detection of edge information in a character region in an MATLAB environment; secondly, filling the detected closed edge, and finishing the selection of the double-mark points according to the area of the connected domain of the double-mark points and the distance of the central point of the connected domain of the double-mark points; and finally, constructing a vector A by taking the square mark point as a starting point and the circular mark point as an end point, constructing a horizontal vector B by passing through the middle point of the connecting line of the two mark points, calculating an included angle theta between the vector A and the vector B from left to right, rotating the character image by the angle theta to obtain a corrected image, wherein the area S of a connected domain and the distance d between the central points of the connected domain of the two mark points need to satisfy the following conditions: d ═ max (d)ij)、20＜S<80, wherein dijThe distance between any two different connected domains is represented, and i and j are the numbers of the two connected domains respectively;
(2) based on projection method character segmentation, the method comprises the following specific steps: firstly, calculating horizontal projection of a character area and finishing horizontal segmentation of the character area; then, calculating the vertical projection of the character area image; finally, according to the width W of the mark pointSign boardRange, Width Range W of character "11The width ranges W of the characters "0", "2", "3", "4", "5", "6", "7", "8", and 9OthersCompleting vertical segmentation of the character region, and segmenting the character region into single characters, wherein W is more than or equal to 6Sign board≤10、9≤W1≤11、17≤WOthers≤23。
6. The method for recognizing the end face characters of the rods in the bundle according to claim 1, wherein in the step 5, a character recognition method based on an SVM classifier is adopted to complete character recognition, and the specific steps comprise:
(1) training an SVM classifier, normalizing character images in an MATLAB environment, finishing training of the SVM classifier through a fitcecac function, and storing a file which is finished by training by using a save function for subsequent identification;
(2) and (3) using an SVM classifier to recognize characters, normalizing the character image to be recognized, and using a predict function to complete character recognition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110657615.1A CN113378846A (en) | 2021-06-13 | 2021-06-13 | Bundled bar end face character recognition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110657615.1A CN113378846A (en) | 2021-06-13 | 2021-06-13 | Bundled bar end face character recognition method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113378846A true CN113378846A (en) | 2021-09-10 |
Family
ID=77574361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110657615.1A Pending CN113378846A (en) | 2021-06-13 | 2021-06-13 | Bundled bar end face character recognition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113378846A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106293779A (en) * | 2015-05-11 | 2017-01-04 | 广州爱九游信息技术有限公司 | A kind of method and device identifying application installation kit |
| CN109840522A (en) * | 2019-01-08 | 2019-06-04 | 河北科技大学 | A kind of big plane rod end tagging scheme and character picture antidote |
| CN110490207A (en) * | 2019-08-27 | 2019-11-22 | 河北科技大学 | Bar section character picture recognition methods based on bianry image gray level co-occurrence matrixes |
| CN110503096A (en) * | 2019-08-27 | 2019-11-26 | 河北科技大学 | The bar section tagging scheme and character picture antidote of horizontal double-flag point |
| CN112496039A (en) * | 2020-11-09 | 2021-03-16 | 广东韶钢松山股份有限公司 | Method for prolonging service life of bar segmentation rolling finished product hole |
-
2021
- 2021-06-13 CN CN202110657615.1A patent/CN113378846A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106293779A (en) * | 2015-05-11 | 2017-01-04 | 广州爱九游信息技术有限公司 | A kind of method and device identifying application installation kit |
| CN109840522A (en) * | 2019-01-08 | 2019-06-04 | 河北科技大学 | A kind of big plane rod end tagging scheme and character picture antidote |
| CN110490207A (en) * | 2019-08-27 | 2019-11-22 | 河北科技大学 | Bar section character picture recognition methods based on bianry image gray level co-occurrence matrixes |
| CN110503096A (en) * | 2019-08-27 | 2019-11-26 | 河北科技大学 | The bar section tagging scheme and character picture antidote of horizontal double-flag point |
| CN112496039A (en) * | 2020-11-09 | 2021-03-16 | 广东韶钢松山股份有限公司 | Method for prolonging service life of bar segmentation rolling finished product hole |
Non-Patent Citations (4)
| Title |
|---|
| 刘子豪: "圆钢端面喷漆机器人喷涂系统关键技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 孔珍: "基于DSP的多种图像增强功能的实现", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 张付祥: "特钢棒材标记方案设计与信息码识别", 《中国冶金》 * |
| 赵阳: "钢坯机器人标识系统关键技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106096610B (en) | A kind of file and picture binary coding method based on support vector machines | |
| CN104573674B (en) | Towards the bar code recognition methods of real time embedded system | |
| CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
| CN112308854B (en) | Automatic detection method and system for chip surface flaws and electronic equipment | |
| CN105632015B (en) | A kind of bank money fingerprint characteristic anti false authentication method based on cloud platform | |
| CN104794502A (en) | Image processing and mode recognition technology-based rice blast spore microscopic image recognition method | |
| CN113034488B (en) | Visual inspection method for ink-jet printed matter | |
| CN116704516B (en) | Visual inspection method for water-soluble fertilizer package | |
| CN113793337A (en) | Locomotive accessory surface abnormal degree evaluation method based on artificial intelligence | |
| CN109781737A (en) | A kind of detection method and its detection system of hose surface defect | |
| CN117094975A (en) | Method and device for detecting surface defects of steel and electronic equipment | |
| CN110276759A (en) | A kind of bad line defect diagnostic method of Mobile phone screen based on machine vision | |
| Vishwanath et al. | Connected component analysis for Indian license plate infra-red and color image character segmentation | |
| CN115078365A (en) | Soft package printing quality defect detection method | |
| CN109741273A (en) | A kind of mobile phone photograph low-quality images automatically process and methods of marking | |
| CN113392819B (en) | Batch academic image automatic segmentation and labeling device and method | |
| CN107145888A (en) | Video caption real time translating method | |
| CN113223098B (en) | Preprocessing optimization method for image color classification | |
| Bala et al. | Image simulation for automatic license plate recognition | |
| CN117911419A (en) | Method and device for detecting steel rotation angle enhancement of medium plate, medium and equipment | |
| CN112686872A (en) | Wood counting method based on deep learning | |
| CN113378846A (en) | Bundled bar end face character recognition method | |
| CN106056575A (en) | Image matching method based on object similarity recommended algorithm | |
| CN108734158B (en) | Real-time train number identification method and device | |
| JP4492258B2 (en) | Character and figure recognition and inspection methods |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210910 |
|
| RJ01 | Rejection of invention patent application after publication |