CN110197181A - A kind of cable character detection method and system based on OCR - Google Patents
A kind of cable character detection method and system based on OCR Download PDFInfo
- Publication number
- CN110197181A CN110197181A CN201910473296.1A CN201910473296A CN110197181A CN 110197181 A CN110197181 A CN 110197181A CN 201910473296 A CN201910473296 A CN 201910473296A CN 110197181 A CN110197181 A CN 110197181A
- Authority
- CN
- China
- Prior art keywords
- character
- cable
- template
- detected
- classifier
- 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.)
- Granted
Links
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
- 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
- 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)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Character Discrimination (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of cable character detection method and system based on OCR is related to image detection identification field, and this method includes the character obtained on normal cable jacket, using each section of character as an element；All elements are linked in sequence as doubly linked list, generate template based on the character in the doubly linked list；Classifier is trained using the template as training sample；The classifier that character input training on cable to be detected is completed, identification is compared with the character in template in character on cable to be detected by classifier, if can recognize that the character on cable to be detected, then show character zero defect on cable to be detected, conversely, then character existing defects on cable to be detected.The present invention, which can fast implement printed character on cable, whether there is the detection of defect, instead of artificial detection, reduce production cost.
Description
Technical field
The present invention relates to image detections to identify field, and in particular to a kind of cable character detection method based on OCR and is
System.
Background technique
In optical cable production process, character apparent for sheath, when lettering block Rig up error or when imprinting shallow, character
Show it is unintelligible, partial character missing, show so as to cause information such as optical cable model, length it is imperfect, influence the later period lay apply
Work, customer experience are impaired；When inker coining is too deep, cable core sleeve stress causes optical fiber attenuation abnormal, seriously affects product
Quality.
Existing character machining mode is artificial detection, and the character matter of identification jacket surface is visually observed by operator
Amount, when cable jacket speed of production is very fast, operator detects speed and is difficult to keep up with, and optical cable production method is continuity
, the case where causing fatigue, being easy to appear missing inspection is persistently detected, meanwhile, the subjective consciousness of different operation personnel is different, detection
Standard be difficult to unification, cause the production efficiency of optical cable to be restricted, product quality is difficult to get a promotion, and cost of labor it is continuous on
Rise.
Summary of the invention
In view of the deficiencies in the prior art, the cable character inspection based on OCR that the purpose of the present invention is to provide a kind of
Method and system are surveyed, can fast implement printed character on cable whether there is the detection of defect, instead of artificial detection, reduce life
Produce cost.
To achieve the above objectives, the technical solution adopted by the present invention is that, comprising:
The character on normal cable jacket is obtained, using each section of character as an element；
All elements are linked in sequence as doubly linked list, generate template based on the character in the doubly linked list；
Classifier is trained using the template as training sample；
The classifier that character input training on cable to be detected is completed, classifier is by the character and mould on cable to be detected
Identification is compared in character in plate, if can recognize that the character on cable to be detected, shows character on cable to be detected
Zero defect, conversely, then character existing defects on cable to be detected.
Based on the above technical solution, the character obtained on normal cable jacket, using each section of character as
One element, specific steps include:
The image of the jacket surface of normal cable is obtained, and binaryzation is carried out to described image；
Based on the mode for obtaining connected domain, character region in image is extracted；
The character in the region is obtained, each section of character is merged according to setting spacing, and each section of character conduct
One element.
Based on the above technical solution,
The image of the normal cable jacket surface of the acquisition is multiframe, and the element in every frame image is connected as one pair
To chained list；
The character based in the doubly linked list generates template, specific steps are as follows: the phase based on each section of intercharacter
Like degree, the corresponding doubly linked list of every frame image is spliced, comprising on normal cable jacket in the doubly linked list spliced
All characters, the character being then based in the obtained doubly linked list of splicing generates template.
Based on the above technical solution, described that classifier is trained using the template as training sample, tool
Body are as follows: the size of the template is normalized, the template after normalized inputs classifier as training sample,
Classifier is trained using SVM algorithm.
Based on the above technical solution, the classifier by the character in the character and template on cable to be detected into
Row matching identification, specific steps include:
Character army in character and template on cable to be detected is zoomed into fixed dimension；
Based on mean square deviation algorithm, each character is calculated on cable to be detected at a distance from intercharacter each in template；
Distance matrix is established, calculates on cable to be detected in every section of character and template every section of intercharacter using DTW algorithm
Distance d chooses the smallest match sequence of all distance d sums.
Based on the above technical solution, described to be based on mean square deviation algorithm, calculate every section of intercharacter on cable to be detected
Distance and template in every section of intercharacter distance, calculation formula are as follows:
Wherein, R (x, y) indicates character pitch from T (x, y) indicates that character position in template, I (x, y) indicate survey line to be checked
Character position on cable.
The cable character machining system based on OCR that the present invention also provides a kind of, comprising:
Module is obtained, is used to obtain the character on normal cable jacket, using each section of character as an element；
Generation module is used to for all elements being linked in sequence as doubly linked list, based on the word in the doubly linked list
Symbol generates template；
Training module is used to be trained classifier using the template as training sample；
Comparison module is used for the classifier for completing the character input training on cable to be detected, and classifier will be to be checked
Identification is compared with the character in template in character on survey line cable, if can recognize that the character on cable to be detected, table
Character zero defect on bright cable to be detected, conversely, then character existing defects on cable to be detected.
Based on the above technical solution, the character obtained on normal cable jacket, using each section of character as
One element, specific steps include:
The image of the jacket surface of normal cable is obtained, and binaryzation is carried out to described image；
Based on the mode for obtaining connected domain, character region in image is extracted；
The character in the region is obtained, each section of character is merged according to setting spacing, and each section of character conduct
One element.
Based on the above technical solution,
The image of the normal cable jacket surface of the acquisition is multiframe, and the element in every frame image is connected as one pair
To chained list；
The character based in the doubly linked list generates template, specific steps are as follows: the phase based on each section of intercharacter
Like degree, the corresponding doubly linked list of every frame image is spliced, comprising on normal cable jacket in the doubly linked list spliced
All characters, the character being then based in the obtained doubly linked list of splicing generates template.
Based on the above technical solution, described that classifier is trained using the template as training sample, tool
Body are as follows: the size of the template is normalized, the template after normalized inputs classifier as training sample,
Classifier is trained using SVM algorithm.
Compared with the prior art, the advantages of the present invention are as follows: by obtaining the character on normal cable jacket, it is connected as double
To chained list, the character being then based in doubly linked list generates template, is trained using template as training sample to classifier, into
And treat the character on detection cable using the classifier after the completion of training and identified, it is quickly real using image recognition technology
Printed character whether there is the detection of defect on existing cable, instead of artificial detection, reduce production cost.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the cable character detection method based on OCR in the embodiment of the present invention；
Fig. 2 is the exemplary diagram of character on cable jacket in the embodiment of the present invention；
Fig. 3 is string matching mode figure in the embodiment of the present invention；
Fig. 4 is DTW matching matrix figure in the embodiment of the present invention.
Specific embodiment
The embodiment of the present invention provides a kind of based on OCR (Optical Character Recognition, optical character knowledge
Cable character detection method not) is based on image recognition technology, and fast implementing printed character on cable whether there is the inspection of defect
It surveys, instead of artificial detection, reduces production cost, improve production efficiency.The embodiment of the present invention has also correspondingly provided one kind and has been based on
The cable character machining system of OCR.
Shown in Figure 1, a kind of cable character detection method based on OCR provided in an embodiment of the present invention includes:
S1: obtaining the character on normal cable jacket, using each section of character as an element.Character is English and number
Form, the information such as specification, production for indicating cable.Fig. 2 example, word may refer to for the pattern of character on cable jacket
Symbol is present on sheath in the form of paragraph, and each section of character includes at least two character, wherein " 2008 " are one section of character,
" FIBERHOME " is one section of character, and " CFOA-SM-AS80-S " is one section of character.For the cable that model determines, on sheath
Character content it is substantially changeless, a complete cable, the character on sheath consists of two parts, fixed word
Symbol part and rice number part, fixed character part immobilize, and fixed character part is by multistage for indicating the information such as model
Character composition, rice number part changes with the variation of length of cable, belongs to variable part, only to sheath in the embodiment of the present invention
On fixed character part carry out recognition detection, rice number part does not identify.
S2: all elements are linked in sequence as doubly linked list, generate template based on the character in the doubly linked list.
In the embodiment of the present invention, since cable is that strip can not be got when character obtains on carrying out cable jacket
All characters present on sheath, can only obtain one section, therefore need to obtain the multistage character of sheath in operation, using duplicate removal
Mode, all characters that will acquire are spliced as unit of element, so that comprising all on sheath in the template generated
Character field, i.e., character field present on normal cable include in template.
S3: classifier is trained using the template as training sample.The step specifically: to the ruler of the template
Very little to be normalized, the template after normalized inputs classifier as training sample, using SVM (Support
Vector Machine, support vector machines) algorithm is trained classifier.It, can be by template when being normalized
Size normalize to length and width be 32 image on, and be divided into 36 classes, wherein 34 classes are number 0~9 and removal " I " and " O "
Letter, remaining 2 classes are used to store the additional character "=" and negative example in rice number region.
S4: the classifier that the character input training on cable to be detected is completed, classifier is by the character on cable to be detected
Identification is compared with the character in template, if can recognize that the character on cable to be detected, shows on cable to be detected
Character zero defect, conversely, then character existing defects on cable to be detected.
Due to including all complete character sections on sheath in template, classifier is by the character and template on cable to be detected
In character identification is compared, if can recognize that the character shown the character on cable to be detected on cable to be detected
It is present in template, shows character zero defect on cable to be detected.
The cable character detection method based on OCR of the embodiment of the present invention, by obtaining the character on normal cable jacket,
Be connected as doubly linked list, be then based on character in doubly linked list and generate template, using template as training sample to classifier into
Row training, and then treat the character on detection cable using the classifier after the completion of training and identified, using image recognition skill
Art, fast implementing printed character on cable whether there is the detection of defect, instead of artificial detection, reduce production cost.
Optionally, on the basis of above-mentioned Fig. 1 corresponding embodiment, a kind of line based on OCR provided in an embodiment of the present invention
In first alternative embodiment of cable character detection method, obtain the character on normal cable jacket, using each section of character as
One element, specific steps include:
S101: the image of the jacket surface of normal cable is obtained, and binaryzation is carried out to described image；
S102: based on the mode for obtaining connected domain, character region in image is extracted；
S103: the character in the region is obtained, each section of character is merged according to setting spacing, and each section of character
As an element.
The image of the normal cable jacket surface obtained is multiframe, and the element in every frame image is connected as a Two-way Chain
Table.
Template, specific steps are generated based on the character in the doubly linked list are as follows: the similarity based on each section of intercharacter,
The corresponding doubly linked list of every frame image is spliced, comprising all on normal cable jacket in the doubly linked list spliced
Character, the character being then based in the doubly linked list that splicing obtains generate template.
Splicing for character field between each doubly linked list, similarity can pass through DTW (dynamic time rule between character field
It is whole) algorithm calculating, specific calculating process is illustrated below, the matching characteristic of character field indicated in a manner of BA, such as
The character representation of " 12FO " is B12F0A1A2AFA0, by taking judgement " TS " and " 12F0 " as an example, string matching mode figure such as Fig. 3 institute
Show, " B is found by DTW algorithm12ATAS" and " B12F0A1A2AFA0" the shortest distance, calculation can be converted into shortest path calculation
Method, establishes distance matrix, and matching result value is shown as shown in Figure 4.
Optionally, on the basis of a kind of above-mentioned first alternative embodiment of cable character detection method based on OCR, this
In a kind of second alternative embodiment of cable character detection method based on OCR that inventive embodiments provide, classifier will be to
Identification is compared with the character in template in character on detection cable, and specific steps include:
Character army in character and template on cable to be detected is zoomed into fixed dimension；
Based on mean square deviation algorithm, each character is calculated on cable to be detected at a distance from intercharacter each in template；
Distance matrix is established, calculates on cable to be detected in every section of character and template every section of intercharacter using DTW algorithm
Distance d chooses the smallest match sequence of all distance d sums.
Based on mean square deviation algorithm, every section of intercharacter in the distance and template of every section of intercharacter is calculated on cable to be detected
Distance, calculation formula are as follows:
Wherein, R (x, y) indicates character pitch from T (x, y) indicates that character position in template, I (x, y) indicate survey line to be checked
Character position on cable.
Due to that can not confirm that the character field of cable to be detected whether there is missing and deletion sites at which, it is therefore desirable to
Character field sequence on cable to be detected and the character field sequence in template are done into constrained arrangement, then from all arrangement L
One sequence o ' of middle extraction, so that sequence distance f (d) is minimum, calculation formula are as follows:
Wherein, Lo'Indicate the sequence taken out, L' indicates standard sequence.
A kind of cable character machining system based on OCR provided in an embodiment of the present invention, comprising:
Module is obtained, is used to obtain the character on normal cable jacket, using each section of character as an element；
Generation module is used to for all elements being linked in sequence as doubly linked list, based on the word in the doubly linked list
Symbol generates template；
Training module is used to be trained classifier using the template as training sample；
Comparison module is used for the classifier for completing the character input training on cable to be detected, and classifier will be to be checked
Identification is compared with the character in template in character on survey line cable, if can recognize that the character on cable to be detected, table
Character zero defect on bright cable to be detected, conversely, then character existing defects on cable to be detected.
The character on normal cable jacket is obtained, using each section of character as an element, specific steps include:
The image of the jacket surface of normal cable is obtained, and binaryzation is carried out to described image；
Based on the mode for obtaining connected domain, character region in image is extracted；
The character in the region is obtained, each section of character is merged according to setting spacing, and each section of character conduct
One element.
The image of the normal cable jacket surface obtained is multiframe, and the element in every frame image is connected as a Two-way Chain
Table；
Template, specific steps are generated based on the character in the doubly linked list are as follows: the similarity based on each section of intercharacter,
The corresponding doubly linked list of every frame image is spliced, comprising all on normal cable jacket in the doubly linked list spliced
Character, the character being then based in the doubly linked list that splicing obtains generate template.Using the template as training sample to classification
Device is trained, specifically: the size of the template is normalized, the template after normalized is as training sample
This input classifier is trained classifier using SVM algorithm.
The cable character machining system based on OCR of the embodiment of the present invention, by obtaining the character on normal cable jacket,
Be connected as doubly linked list, be then based on character in doubly linked list and generate template, using template as training sample to classifier into
Row training, and then treat the character on detection cable using the classifier after the completion of training and identified, using image recognition skill
Art, fast implementing printed character on cable whether there is the detection of defect, instead of artificial detection, reduce production cost.
The present invention is not limited to the above-described embodiments, for those skilled in the art, is not departing from
Under the premise of the principle of the invention, several improvements and modifications can also be made, these improvements and modifications are also considered as protection of the invention
Within the scope of.The content being not described in detail in this specification belongs to the prior art well known to professional and technical personnel in the field.
Claims (10)
1. a kind of cable character detection method based on OCR, which comprises the following steps:
The character on normal cable jacket is obtained, using each section of character as an element；
All elements are linked in sequence as doubly linked list, generate template based on the character in the doubly linked list；
Classifier is trained using the template as training sample；
The classifier that character input training on cable to be detected is completed, classifier will be in the character and template on cable to be detected
Character identification is compared, if can recognize that the character on cable to be detected, show that character is intact on cable to be detected
It falls into, conversely, then character existing defects on cable to be detected.
2. a kind of cable character detection method based on OCR as described in claim 1, which is characterized in that described to obtain normally
Character on cable jacket, using each section of character as an element, specific steps include:
The image of the jacket surface of normal cable is obtained, and binaryzation is carried out to described image；
Based on the mode for obtaining connected domain, character region in image is extracted；
The character in the region is obtained, each section of character is merged according to setting spacing, and each section of character is as one
Element.
3. a kind of cable character detection method based on OCR as claimed in claim 2, it is characterised in that:
The image of the normal cable jacket surface of the acquisition is multiframe, and the element in every frame image is connected as a Two-way Chain
Table；
The character based in the doubly linked list generates template, specific steps are as follows: the similarity based on each section of intercharacter,
The corresponding doubly linked list of every frame image is spliced, comprising all on normal cable jacket in the doubly linked list spliced
Character, the character being then based in the doubly linked list that splicing obtains generate template.
4. a kind of cable character detection method based on OCR as described in claim 1, which is characterized in that described by the mould
Plate is trained classifier as training sample, specifically: the size of the template is normalized, at normalization
Template after reason inputs classifier as training sample, is trained using SVM algorithm to classifier.
5. a kind of cable character detection method based on OCR as described in claim 1, which is characterized in that the classifier will
Identification is compared with the character in template in character on cable to be detected, and specific steps include:
Character army in character and template on cable to be detected is zoomed into fixed dimension；
Based on mean square deviation algorithm, each character is calculated on cable to be detected at a distance from intercharacter each in template；
Establish distance matrix, using DTW algorithm calculate on cable to be detected every section of character in template at a distance from every section of intercharacter
D chooses the smallest match sequence of all distance d sums.
6. a kind of cable character detection method based on OCR as claimed in claim 5, which is characterized in that described based on square
Difference algorithm calculates the distance of every section of intercharacter in the distance and template of every section of intercharacter on cable to be detected, calculation formula
Are as follows:
Wherein, R (x, y) indicates character pitch from T (x, y) indicates that character position in template, I (x, y) indicate on cable to be detected
Character position.
7. a kind of cable character machining system based on OCR characterized by comprising
Module is obtained, is used to obtain the character on normal cable jacket, using each section of character as an element；
Generation module is used to for all elements being linked in sequence as doubly linked list, raw based on the character in the doubly linked list
At template；
Training module is used to be trained classifier using the template as training sample；
Comparison module is used for the classifier for completing the character input training on cable to be detected, and classifier is by survey line to be checked
Identification is compared with the character in template in character on cable, if can recognize that the character on cable to be detected, show to
Character zero defect on cable is detected, conversely, then character existing defects on cable to be detected.
8. a kind of cable character machining system based on OCR as claimed in claim 7, which is characterized in that described to obtain normally
Character on cable jacket, using each section of character as an element, specific steps include:
The image of the jacket surface of normal cable is obtained, and binaryzation is carried out to described image；
Based on the mode for obtaining connected domain, character region in image is extracted；
The character in the region is obtained, each section of character is merged according to setting spacing, and each section of character is as one
Element.
9. a kind of cable character machining system based on OCR as claimed in claim 8, it is characterised in that:
The image of the normal cable jacket surface of the acquisition is multiframe, and the element in every frame image is connected as a Two-way Chain
Table；
The character based in the doubly linked list generates template, specific steps are as follows: the similarity based on each section of intercharacter,
The corresponding doubly linked list of every frame image is spliced, comprising all on normal cable jacket in the doubly linked list spliced
Character, the character being then based in the doubly linked list that splicing obtains generate template.
10. a kind of cable character machining system based on OCR as claimed in claim 7, which is characterized in that described by the mould
Plate is trained classifier as training sample, specifically: the size of the template is normalized, at normalization
Template after reason inputs classifier as training sample, is trained using SVM algorithm to classifier.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910473296.1A CN110197181B (en) | 2019-05-31 | 2019-05-31 | Cable character detection method and system based on OCR |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910473296.1A CN110197181B (en) | 2019-05-31 | 2019-05-31 | Cable character detection method and system based on OCR |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110197181A true CN110197181A (en) | 2019-09-03 |
| CN110197181B CN110197181B (en) | 2021-04-30 |
Family
ID=67753714
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910473296.1A Active CN110197181B (en) | 2019-05-31 | 2019-05-31 | Cable character detection method and system based on OCR |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110197181B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114780685A (en) * | 2022-04-28 | 2022-07-22 | 贵州电网有限责任公司 | Method for automatically identifying defect information input condition and supplementing defect information through unmanned aerial vehicle |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130044149A1 (en) * | 2011-08-19 | 2013-02-21 | Takuya Moriai | Inkjet recording apparatus |
| CN103488986A (en) * | 2013-09-18 | 2014-01-01 | 西安理工大学 | Method for segmenting and extracting characters in self-adaptation mode |
| CN104820986A (en) * | 2015-04-28 | 2015-08-05 | 电子科技大学 | Machine vision-based cable on-line detection method |
| CN105260475A (en) * | 2015-10-30 | 2016-01-20 | 努比亚技术有限公司 | Data searching method, data saving method and related equipment |
| CN106227668A (en) * | 2016-07-29 | 2016-12-14 | 腾讯科技（深圳）有限公司 | Data processing method and device |
| CN106650721A (en) * | 2016-12-28 | 2017-05-10 | 吴晓军 | Industrial character identification method based on convolution neural network |
| CN108154144A (en) * | 2018-01-12 | 2018-06-12 | 江苏省新通智能交通科技发展有限公司 | A kind of name of vessel character locating method and system based on image |
| CN108596173A (en) * | 2018-04-19 | 2018-09-28 | 长春理工大学 | One camera full view wire size real-time distinguishing apparatus and its detection method |
| CN109409272A (en) * | 2018-10-17 | 2019-03-01 | 北京空间技术研制试验中心 | Cable Acceptance Test System and method based on machine vision |
| CN109712162A (en) * | 2019-01-18 | 2019-05-03 | 珠海博明视觉科技有限公司 | A kind of cable character defect inspection method and device based on projection histogram difference |
-
2019
- 2019-05-31 CN CN201910473296.1A patent/CN110197181B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130044149A1 (en) * | 2011-08-19 | 2013-02-21 | Takuya Moriai | Inkjet recording apparatus |
| CN103488986A (en) * | 2013-09-18 | 2014-01-01 | 西安理工大学 | Method for segmenting and extracting characters in self-adaptation mode |
| CN104820986A (en) * | 2015-04-28 | 2015-08-05 | 电子科技大学 | Machine vision-based cable on-line detection method |
| CN105260475A (en) * | 2015-10-30 | 2016-01-20 | 努比亚技术有限公司 | Data searching method, data saving method and related equipment |
| CN106227668A (en) * | 2016-07-29 | 2016-12-14 | 腾讯科技（深圳）有限公司 | Data processing method and device |
| CN106650721A (en) * | 2016-12-28 | 2017-05-10 | 吴晓军 | Industrial character identification method based on convolution neural network |
| CN108154144A (en) * | 2018-01-12 | 2018-06-12 | 江苏省新通智能交通科技发展有限公司 | A kind of name of vessel character locating method and system based on image |
| CN108596173A (en) * | 2018-04-19 | 2018-09-28 | 长春理工大学 | One camera full view wire size real-time distinguishing apparatus and its detection method |
| CN109409272A (en) * | 2018-10-17 | 2019-03-01 | 北京空间技术研制试验中心 | Cable Acceptance Test System and method based on machine vision |
| CN109712162A (en) * | 2019-01-18 | 2019-05-03 | 珠海博明视觉科技有限公司 | A kind of cable character defect inspection method and device based on projection histogram difference |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114780685A (en) * | 2022-04-28 | 2022-07-22 | 贵州电网有限责任公司 | Method for automatically identifying defect information input condition and supplementing defect information through unmanned aerial vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110197181B (en) | 2021-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106934800B (en) | Metal plate strip surface defect detection method and device based on YOLO9000 network | |
| CN106204618A (en) | Product surface of package defects detection based on machine vision and sorting technique | |
| CN108492291B (en) | CNN segmentation-based solar photovoltaic silicon wafer defect detection system and method | |
| CN108562589A (en) | A method of magnetic circuit material surface defect is detected | |
| CN108763931A (en) | Leak detection method based on Bi-LSTM and text similarity | |
| CN105260734A (en) | Commercial oil surface laser code recognition method with self modeling function | |
| CN105930836A (en) | Identification method and device of video text | |
| CN105389558A (en) | Method and apparatus for detecting video | |
| CN113083804A (en) | Laser intelligent derusting method and system and readable medium | |
| CN109839386B (en) | Intelligent camera shooting identification system | |
| CN111899216A (en) | Abnormity detection method for insulator fastener of high-speed rail contact network | |
| CN114723708A (en) | Handicraft appearance defect detection method based on unsupervised image segmentation | |
| CN108681538A (en) | A kind of verb phrase omission digestion procedure based on deep learning | |
| CN114494780A (en) | Semi-supervised industrial defect detection method and system based on feature comparison | |
| CN112907562A (en) | MobileNet-based SMT defect classification algorithm | |
| CN110197181A (en) | A kind of cable character detection method and system based on OCR | |
| CN105975802A (en) | Grading method and device for CAD drawing | |
| CN115019294A (en) | Pointer instrument reading identification method and system | |
| CN111582331A (en) | Painting work author image identification method based on convolutional neural network | |
| CN110689447A (en) | Real-time detection method for social software user published content based on deep learning | |
| CN111105396A (en) | Printed matter quality detection method and system based on artificial intelligence | |
| CN110196897A (en) | A kind of case recognition methods based on question and answer template | |
| CN116755647B (en) | Printing method and system for preventing printing data errors | |
| US11132572B2 (en) | Method and system for splicing and restoring shredded paper based on extreme learning machine | |
| CN114926702B (en) | Small sample image classification method based on depth attention measurement |
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 |