CN107944500A - The similar decision method of picture that a kind of HOG is combined with histogram - Google Patents
The similar decision method of picture that a kind of HOG is combined with histogram Download PDFInfo
- Publication number
- CN107944500A CN107944500A CN201711305448.4A CN201711305448A CN107944500A CN 107944500 A CN107944500 A CN 107944500A CN 201711305448 A CN201711305448 A CN 201711305448A CN 107944500 A CN107944500 A CN 107944500A
- Authority
- CN
- China
- Prior art keywords
- histogram
- gradient
- picture
- hashed
- value
- 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.)
- Withdrawn
Links
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
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/758—Involving statistics of pixels or of feature values, e.g. histogram matching
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses the similar decision method of picture that a kind of HOG is combined with histogram, between accurate judgement, adds segmentation gradient magnitude：Gradient magnitude is divided into 4*4 regions, calculates the gradient magnitude histogram value in each region；With compare histogram hashed value：Histogram value uniformity is hashed, obtains histogram hashed value, compares the histogram hashed value of two pictures, more than or equal to 5, is then judged similar；If similar pictures quantity is more than the step of threshold value of setting.The present invention is by the way that before all gray-scale pixels hash, the histogram for doing a piecemeal hashes, and reduces N times of calculation amount；N is the number of piecemeal；Rough judgement is first provided, then does and further precisely judges again.
Description
Technical field
The present invention relates to image processing field, the similar decision method of picture that more particularly to a kind of HOG is combined with histogram.
Background technology
Applicant is in patent application entitled " one kind is based on the similar decision method of the improved pictures of HOG ", in advance to figure
Piece is compressed processing, then calculates gradient, and the feature extraction fixed to gradient also uses different methods, is not division CELL
Mode, but calculate average after contrast, formed hash, accelerate calculating speed.
The above method is direct generation judging result, and in practical applications, more than 80% picture is without accurately sentence very much
It is disconnected, you can to exclude approximately may not；So before precise results are obtained, add a simple anticipation process, can be into one
Step improves computational efficiency.
The content of the invention
In order to overcome the above-mentioned deficiencies of the prior art, the picture combined with histogram the present invention provides a kind of HOG is similar
Decision method.
The similar decision method of picture that a kind of HOG provided by the invention is combined with histogram, comprises the following steps：
1) picture ratio is reduced
Picture is reduced to 32*32 sizes；
2) gray processing
The pixel of picture is down to 32 grades of gray scales；Grey level is consistent with dimension of picture；That is first step drawdown ratio
If example is 64*64, then pixel scale 64；
3) gradient is calculated
Calculate the gradient of each pixel of image, including size and Orientation；
4) gradient magnitude is split
Gradient magnitude is divided into 4*4 regions, calculates the gradient magnitude histogram value in each region；
5) histogram hashed value is compared
Histogram value uniformity is hashed, obtains histogram hashed value, compares the histogram hashed value of two pictures, is more than
Or equal to 5, then judge similar；If similar pictures quantity is more than the threshold value of setting；Only carry out the 6) step；Further screening；
6) gradient magnitude average is calculated
7) gradient magnitude is compared
The gradient magnitude of each pixel is compared with gradient magnitude average, more than or equal to for 1, less than for 0；
8) hash
First pixel gradient size comparative result uniformity is hashed, then hashes pixel orientation uniformity, forms one
Integer gradient hashed value；
9) gradient hashed value is compared
Different gradient hashed values are compared, Hamming distance is less than or equal to 5, then judges that picture is similar or identical.
Beneficial effect：The present invention is by the way that before all gray-scale pixels hash, the histogram for doing a piecemeal hashes, and reduces
N times of calculation amount；N is the number of piecemeal；Rough judgement is first provided, then does and further precisely judges again.
Embodiment
Embodiment：
1) picture ratio is reduced
Picture is reduced to 32*32 sizes；
2) gray processing
The pixel of picture is down to 32 grades of gray scales；Grey level is consistent with dimension of picture；That is first step drawdown ratio
If example is 64*64, then pixel scale 64；
3) gradient is calculated
Calculate the gradient of each pixel of image, including size and Orientation；
4) gradient magnitude is split
Gradient magnitude is divided into 4*4 regions, calculates the gradient magnitude histogram value in each region；
5) histogram hashed value is compared
Histogram value uniformity is hashed, obtains histogram hashed value, compares the histogram hashed value of two pictures, is more than
Or equal to 5, then judge similar；If similar pictures quantity is more than the threshold value of setting；Only carry out the 6) step；Further screening；
6) gradient magnitude average is calculated
7) gradient magnitude is compared
The gradient magnitude of each pixel is compared with gradient magnitude average, more than or equal to for 1, less than for 0；
8) hash
First pixel gradient size comparative result uniformity is hashed, then hashes pixel orientation uniformity, forms one
Integer gradient hashed value；
9) gradient hashed value is compared
Different gradient hashed values are compared, Hamming distance is less than or equal to 5, then judges that picture is similar or identical.
Claims (1)
1. the similar decision method of picture that a kind of HOG is combined with histogram, it is characterised in that comprise the following steps：
1) picture ratio is reduced
Picture is reduced to 32*32 sizes；
2) gray processing
The pixel of picture is down to 32 grades of gray scales；Grey level is consistent with dimension of picture；That is first step diminution ratio is
If 64*64, then pixel scale 64；
3) gradient is calculated
Calculate the gradient of each pixel of image, including size and Orientation；
4) gradient magnitude is split
Gradient magnitude is divided into 4*4 regions, calculates the gradient magnitude histogram value in each region；
5) histogram hashed value is compared
Histogram value uniformity is hashed, obtains histogram hashed value, compares the histogram hashed value of two pictures, is more than or waits
In 5, then judge similar；If similar pictures quantity is more than the threshold value of setting；Only carry out the 6) step；Further screening；
6) gradient magnitude average is calculated
7) gradient magnitude is compared
The gradient magnitude of each pixel is compared with gradient magnitude average, more than or equal to for 1, less than for 0；
8) hash
First pixel gradient size comparative result uniformity is hashed, then hashes pixel orientation uniformity, forms an integer
Gradient hashed value；
9) gradient hashed value is compared
Different gradient hashed values are compared, Hamming distance is less than or equal to 5, then judges that picture is similar or identical.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711305448.4A CN107944500A (en) | 2017-12-11 | 2017-12-11 | The similar decision method of picture that a kind of HOG is combined with histogram |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711305448.4A CN107944500A (en) | 2017-12-11 | 2017-12-11 | The similar decision method of picture that a kind of HOG is combined with histogram |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107944500A true CN107944500A (en) | 2018-04-20 |
Family
ID=61945485
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711305448.4A Withdrawn CN107944500A (en) | 2017-12-11 | 2017-12-11 | The similar decision method of picture that a kind of HOG is combined with histogram |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107944500A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105069042A (en) * | 2015-07-23 | 2015-11-18 | 北京航空航天大学 | Content-based data retrieval methods for unmanned aerial vehicle spying images |
| WO2016033676A1 (en) * | 2014-09-02 | 2016-03-10 | Netra Systems Inc. | System and method for analyzing and searching imagery |
| CN106503534A (en) * | 2015-09-08 | 2017-03-15 | 腾讯科技（深圳）有限公司 | A kind of information processing method and terminal |
| CN107145487A (en) * | 2016-03-01 | 2017-09-08 | 深圳中兴力维技术有限公司 | Image search method and device |
-
2017
- 2017-12-11 CN CN201711305448.4A patent/CN107944500A/en not_active Withdrawn
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016033676A1 (en) * | 2014-09-02 | 2016-03-10 | Netra Systems Inc. | System and method for analyzing and searching imagery |
| CN105069042A (en) * | 2015-07-23 | 2015-11-18 | 北京航空航天大学 | Content-based data retrieval methods for unmanned aerial vehicle spying images |
| CN106503534A (en) * | 2015-09-08 | 2017-03-15 | 腾讯科技（深圳）有限公司 | A kind of information processing method and terminal |
| CN107145487A (en) * | 2016-03-01 | 2017-09-08 | 深圳中兴力维技术有限公司 | Image search method and device |
Non-Patent Citations (1)
| Title |
|---|
| 张现波: "手绘草图的图像检索技术研究及***实现", 《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 |
|---|---|---|
| CN103870828A (en) | System and method for judging image similarity degree | |
| US10395374B2 (en) | Surveillance video based video foreground extraction method | |
| JP2018508875A5 (en) | ||
| CN115546203B (en) | Production monitoring and analyzing method based on image data algorithm | |
| CN112508090A (en) | External package defect detection method | |
| CN108133463B (en) | Noise reduction method and system for histogram equalization image | |
| CN102509095B (en) | Number plate image preprocessing method | |
| CN111353552A (en) | Image similarity contrast method based on perceptual hash algorithm | |
| CN100493134C (en) | Method and system for processing image | |
| CN114494318B (en) | Cornea contour extraction method based on cornea dynamic deformation video of Ojin algorithm | |
| CN106991753B (en) | A kind of image binaryzation method and device | |
| CN109978858B (en) | Double-frame thumbnail image quality evaluation method based on foreground detection | |
| US10796442B2 (en) | Two-dimensional image depth of field generation method and device | |
| US20160277657A1 (en) | Video processing method and apparatus | |
| CN107944500A (en) | The similar decision method of picture that a kind of HOG is combined with histogram | |
| CN110705568B (en) | Optimization method for image feature point extraction | |
| CN105069819A (en) | Predicted value compensation method based on MED predication algorithm | |
| CN104657972A (en) | Ambiguity judging method and system of image block | |
| CN114998186B (en) | Method and system for detecting surface scab defect of copper starting sheet based on image processing | |
| CN107784269A (en) | A kind of method and system of 3D frame of video feature point extraction | |
| CN109685114B (en) | Image similarity judgment method based on pre-screening method and PHash | |
| CN108052972A (en) | A kind of single pass picture similar determination method of Two bors d's oeuveres | |
| CN112561949B (en) | Rapid moving object detection algorithm based on RPCA and support vector machine | |
| CN115880181A (en) | Method, device and terminal for enhancing image contrast | |
| CN113674259B (en) | Belt conveyor slip detection method, system, electronic equipment and medium |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180420 |