CN103714122A - Image retrieval method based on local block binary coding characteristics - Google Patents
Image retrieval method based on local block binary coding characteristics Download PDFInfo
- Publication number
- CN103714122A CN103714122A CN201310651718.2A CN201310651718A CN103714122A CN 103714122 A CN103714122 A CN 103714122A CN 201310651718 A CN201310651718 A CN 201310651718A CN 103714122 A CN103714122 A CN 103714122A
- Authority
- CN
- China
- Prior art keywords
- image
- feature
- region
- local
- coding
- 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
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000010606 normalization Methods 0.000 claims abstract description 11
- 230000008859 change Effects 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 abstract description 8
- 230000000007 visual effect Effects 0.000 abstract 3
- 238000010586 diagram Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/5838—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
Landscapes
- Engineering & Computer Science (AREA)
- Library & Information Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention relates to the field of image retrieval, in particular to an image retrieval method based on local block binary coding characteristics, which comprises the steps of a, inputting an image as a query image; b. carrying out format normalization; c. carrying out scale normalization; d. extracting characteristics; e. calculating all local characteristic values; f. obtaining global features, discretizing and normalizing the local feature histogram to obtain a feature histogram H of N bins in common; g. coding the feature histogram H; i. by comparing the Hamming distance between the feature codes of each image and the query image, the visual approximate image can be obtained through a self-set threshold value. The image retrieval method based on the local block binary coding features combines the visual approximate image detection of the global features and the local features, and the visual approximate images with similar results are obtained through the retrieval method, so that the accuracy is high, and the retrieval is convenient and quick.
Description
Technical field
The present invention relates to the field of image retrieval, especially a kind of image search method based on local piecemeal binary coding feature.
Background technology
Current existing two kinds of image search methods, wherein a kind of method is the detection method (global characteristics) based on color histogram: (1) is separated into B, G, R triple channel image by piece image, add up each passage color histogram H1(256bins altogether), after discretize, be N bins histogram H2; (2) H2 is normalized to (object is to make it not be subject to the impact of sum of all pixels) as far as possible obtains histogram H3; (3) H3 is encoded, comparison result.Above-mentioned detection method, its advantage is that principle is simple, and detection speed is fast, and its shortcoming is to have the situation of obvious change to detect to the color of only having of image, only better to image direction or the vicissitudinous situation effect of size; Second method is the detection method (local feature) based on texture: this is that a kind of bottom texture information arranging based on picture structure detects, what add up is microscopic characteristics (as local features such as edge, lines), its advantage is technology maturation, its shortcoming is to calculate too complexity, program realizes comparatively huge, be unfavorable for understanding, and only to almost identical image or that the image effect of offset effect only occurs is better, to image local feature have very large change situation to detect effect poor.
Above-mentioned first method as shown in Figure 1, for a in Fig. 1 and this size of b and color occur slight change situation detect better, as the obvious situation of this color change in b, (although visually image content is approximate) cannot be detected; Second method can have good detection for above-mentioned three kinds of situations, but for detecting by situation about changing in image local, as shown in Figure 2, two width figure are visually approximate, but people's hand motion changes.
Summary of the invention
The technical problem to be solved in the present invention is: in order to overcome the problem of above-mentioned middle existence, provide a kind of image search method based on local piecemeal binary coding feature, it can realize the object of accurate retrieving images.
The technical solution adopted for the present invention to solve the technical problems is: a kind of image search method based on local piecemeal binary coding feature, and concrete grammar is as follows:
A. input piece image as query image;
B. carry out form normalization: image is converted to RGB color space and isolates three channel image of R passage, G passage, B passage;
C. carry out yardstick normalization: change the width of image into fixed size, the height of image is determined by the constant principle of the ratio of width to height of image;
D. feature extraction: characterizing definition is on the region of 3x3, and by its area marking totally 9 regions from A to I, the region that is positioned at center is E, and the size in each region is identical,
Suppose width and be highly respectively W
b, H
bcalculate 9 pixel intensity in region and, and by the pixel intensity in B, D, F, H, A, C, G, I region and respectively with pixel intensity and the comparison in E region, if the pixel intensity in B, D, F, H, A, C, G, I region and>=pixel intensity in E region and, corresponding binary digit is 1, otherwise be 0, using 8 bits as final Image Coding feature;
Suppose the rectangular area brightness that obtains and, feature coding is respectively 0 (1025<1500) 1 (2301>1500) 1 (1672>1500) 0 (990<1500) 0 (956<1500) 0 (845<1500) 1 (1865>1500) 0 (875<1500), 01100010, corresponding decimal number is 98, i.e. local feature value;
E. with fixing horizontal, move rectangular block with vertical step-length, until calculate all local feature values;
F. add up local feature histogram, obtain global characteristics, local feature histogram is carried out to discretize and normalization, obtain the feature histogram H of total N bins;
G. feature histogram H is encoded:
To n=1,2 ... N-1, suppose the histogram value H (n) to n bin, if H (n-1) <H (n), the N-1-n position of final feature is 1, otherwise is 0; Each channel image has N position coding characteristic, so piece image is total to the total 3*N of triple channel position eigenwert;
H. the every piece image in picture library is all carried out to above-mentioned a~g step, can obtain the 3*N position feature coding of every piece image;
I. by the hamming distance between the feature coding between the every piece image of comparison and query image, by the threshold value of certainly establishing, can obtain vision approximate image
The invention has the beneficial effects as follows, a kind of image search method based on local piecemeal binary coding feature of the present invention, the vision approximate image that this image search method combines global characteristics and local feature detects, the N dimensional feature vector with same procedure obtains of the N dimensional feature vector obtaining according to query image and each pictures in picture library compares, by this search method, result is similar is vision approximate image, and accuracy rate is high, retrieves convenient and swift.
Accompanying drawing explanation
Below in conjunction with drawings and Examples, the present invention is further described.
Fig. 1 is the schematic diagram using in the detection method based on color histogram in background technology of the present invention;
Fig. 2 is the schematic diagram using in the detection method based on texture in background technology of the present invention;
Fig. 3 be in the feature extraction of a kind of image search method based on local piecemeal binary coding feature of the present invention characterizing definition at the schematic diagram in the region of 3x3;
Fig. 4 be in the feature extraction of a kind of image search method based on local piecemeal binary coding feature of the present invention rectangular area brightness and schematic diagram;
Fig. 5 is the schematic diagram of the local feature value of a kind of image search method based on local piecemeal binary coding feature of the present invention;
Fig. 6 is the process flow diagram of a kind of image search method based on local piecemeal binary coding feature of the present invention.
Embodiment
In conjunction with the accompanying drawings, the present invention is further detailed explanation.These accompanying drawings are the schematic diagram of simplification, basic structure of the present invention is only described in a schematic way, so it only show the formation relevant with the present invention.
A kind of image search method based on local piecemeal binary coding feature as shown in Figure 6, concrete grammar is as follows:
A. input piece image as query image;
B. carry out form normalization: image is converted to RGB color space and isolates three channel image of R passage, G passage, B passage;
C. carry out yardstick normalization: change the width of image into fixed size, the height of image is determined by the constant principle of the ratio of width to height of image;
D. feature extraction: characterizing definition is on the region of 3x3, and as shown in Figure 3, by its area marking totally 9 regions from A to I, the region that is positioned at center is E, and the size in each region is identical,
Suppose width and be highly respectively W
b, H
bcalculate 9 pixel intensity in region and, and by the pixel intensity in B, D, F, H, A, C, G, I region and respectively with pixel intensity and the comparison in E region, if the pixel intensity in B, D, F, H, A, C, G, I region and>=pixel intensity in E region and, corresponding binary digit is 1, otherwise be 0, using 8 bits as final Image Coding feature;
Suppose the rectangular area brightness that obtains and, as shown in Figure 4, feature coding is respectively 0 (1025<1500) 1 (2301>1500) 1 (1672>1500) 0 (990<1500) 0 (956<1500) 0 (845<1500) 1 (1865>1500) 0 (875<1500), 01100010, corresponding decimal number is 98, i.e. local feature value;
E. with fixing horizontal, move rectangular block with vertical step-length, until calculate all local feature values, as shown in Figure 5, from the upper left corner, calculate the lower right corner always;
F. add up local feature histogram, obtain global characteristics, local feature histogram is carried out to discretize and normalization, obtain the feature histogram H of total N bins;
G. feature histogram H is encoded:
To n=1,2 ... N-1, suppose the histogram value H (n) to n bin, if H (n-1) <H (n), the N-1-n position of final feature is 1, otherwise is 0; Each channel image has N position coding characteristic, so piece image is total to the total 3*N of triple channel position eigenwert;
H. the every piece image in picture library is all carried out to above-mentioned a~g step, can obtain the 3*N position feature coding of every piece image;
I. by the hamming distance between the feature coding between the every piece image of comparison and query image, by the threshold value of certainly establishing, can obtain vision approximate image.
A kind of image search method based on local piecemeal binary coding feature of the present invention, in test picture library, have 20 groups of images, every group of 4 width images are all vision approximate image, after given query image, there are 18 groups all can accurately detect similar image, other two groups exist undetectedly, and accuracy rate is higher than 90%.
The above-mentioned foundation desirable embodiment of the present invention of take is enlightenment, and by above-mentioned description, relevant staff can, within not departing from the scope of this invention technological thought, carry out various change and modification completely.The technical scope of this invention is not limited to the content on instructions, must determine its technical scope according to claim scope.
Claims (1)
1. the image search method based on local piecemeal binary coding feature, is characterized in that concrete grammar is as follows:
A. input piece image as query image;
B. carry out form normalization: image is converted to RGB color space and isolates three channel image of R passage, G passage, B passage;
C. carry out yardstick normalization: change the width of image into fixed size, the height of image is determined by the constant principle of the ratio of width to height of image;
D. feature extraction: characterizing definition is on the region of 3x3, and by its area marking totally 9 regions from A to I, the region that is positioned at center is E, and the size in each region is identical,
Suppose width and be highly respectively W
b, H
bcalculate 9 pixel intensity in region and, and by the pixel intensity in B, D, F, H, A, C, G, I region and respectively with pixel intensity and the comparison in E region, if the pixel intensity in B, D, F, H, A, C, G, I region and>=pixel intensity in E region and, corresponding binary digit is 1, otherwise be 0, using 8 bits as final Image Coding feature;
Suppose the rectangular area brightness that obtains and, feature coding is respectively 0 (1025<1500) 1 (2301>1500) 1 (1672>1500) 0 (990<1500) 0 (956<1500) 0 (845<1500) 1 (1865>1500) 0 (875<1500), 01100010, corresponding decimal number is 98, i.e. local feature value;
E. with fixing horizontal, move rectangular block with vertical step-length, until calculate all local feature values;
F. add up local feature histogram, obtain global characteristics, local feature histogram is carried out to discretize and normalization, obtain the feature histogram H of total N bins;
G. feature histogram H is encoded:
To n=1,2 ... N-1, suppose the histogram value H (n) to n bin, if H (n-1) <H (n), the N-1-n position of final feature is 1, otherwise is 0; Each channel image has N position coding characteristic, so piece image is total to the total 3*N of triple channel position eigenwert;
H. the every piece image in picture library is all carried out to above-mentioned a~g step, can obtain the 3*N position feature coding of every piece image;
I. by the hamming distance between the feature coding between the every piece image of comparison and query image, by the threshold value of certainly establishing, can obtain vision approximate image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310651718.2A CN103714122A (en) | 2013-12-06 | 2013-12-06 | Image retrieval method based on local block binary coding characteristics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310651718.2A CN103714122A (en) | 2013-12-06 | 2013-12-06 | Image retrieval method based on local block binary coding characteristics |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103714122A true CN103714122A (en) | 2014-04-09 |
Family
ID=50407097
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310651718.2A Pending CN103714122A (en) | 2013-12-06 | 2013-12-06 | Image retrieval method based on local block binary coding characteristics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103714122A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104615612A (en) * | 2014-04-30 | 2015-05-13 | 北京大学 | Method for obtaining code stream of compact global feature descriptors |
CN106354735A (en) * | 2015-07-22 | 2017-01-25 | 杭州海康威视数字技术股份有限公司 | Image target searching method and device |
CN106815589A (en) * | 2015-12-01 | 2017-06-09 | 财团法人工业技术研究院 | Feature description method and feature descriptor using same |
CN108763265A (en) * | 2018-04-03 | 2018-11-06 | 南昌奇眸科技有限公司 | A kind of image-recognizing method based on block research |
CN110251076A (en) * | 2019-06-21 | 2019-09-20 | 安徽大学 | Merge conspicuousness detection method and device of the visual attention based on contrast |
CN111522986A (en) * | 2020-04-23 | 2020-08-11 | 北京百度网讯科技有限公司 | Image retrieval method, apparatus, device and medium |
CN111986089A (en) * | 2020-08-28 | 2020-11-24 | 计易数据科技(上海)有限公司 | Image storage and comparison method, device, equipment and medium with characteristic value being integer |
CN112182262A (en) * | 2020-11-30 | 2021-01-05 | 江西师范大学 | Image query method based on feature classification |
US11132392B2 (en) * | 2019-04-17 | 2021-09-28 | Boe Technology Group Co., Ltd. | Image retrieval method, image retrieval apparatus, image retrieval device and medium |
-
2013
- 2013-12-06 CN CN201310651718.2A patent/CN103714122A/en active Pending
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104615612A (en) * | 2014-04-30 | 2015-05-13 | 北京大学 | Method for obtaining code stream of compact global feature descriptors |
CN104615612B (en) * | 2014-04-30 | 2018-04-27 | 北京大学 | The method for obtaining the code stream of compact global characteristics description |
CN106354735A (en) * | 2015-07-22 | 2017-01-25 | 杭州海康威视数字技术股份有限公司 | Image target searching method and device |
WO2017012277A1 (en) * | 2015-07-22 | 2017-01-26 | 杭州海康威视数字技术股份有限公司 | Method and device for searching a target in an image |
US10678847B2 (en) | 2015-07-22 | 2020-06-09 | Hangzhou Hikvision Digital Technology Co., Ltd. | Method and device for searching a target in an image |
CN106815589A (en) * | 2015-12-01 | 2017-06-09 | 财团法人工业技术研究院 | Feature description method and feature descriptor using same |
CN108763265A (en) * | 2018-04-03 | 2018-11-06 | 南昌奇眸科技有限公司 | A kind of image-recognizing method based on block research |
CN108763265B (en) * | 2018-04-03 | 2022-02-08 | 南昌奇眸科技有限公司 | Image identification method based on block retrieval |
US11132392B2 (en) * | 2019-04-17 | 2021-09-28 | Boe Technology Group Co., Ltd. | Image retrieval method, image retrieval apparatus, image retrieval device and medium |
CN110251076B (en) * | 2019-06-21 | 2021-10-22 | 安徽大学 | Method and device for detecting significance based on contrast and fusing visual attention |
CN110251076A (en) * | 2019-06-21 | 2019-09-20 | 安徽大学 | Merge conspicuousness detection method and device of the visual attention based on contrast |
CN111522986A (en) * | 2020-04-23 | 2020-08-11 | 北京百度网讯科技有限公司 | Image retrieval method, apparatus, device and medium |
CN111522986B (en) * | 2020-04-23 | 2023-10-10 | 北京百度网讯科技有限公司 | Image retrieval method, device, equipment and medium |
CN111986089A (en) * | 2020-08-28 | 2020-11-24 | 计易数据科技(上海)有限公司 | Image storage and comparison method, device, equipment and medium with characteristic value being integer |
CN112182262A (en) * | 2020-11-30 | 2021-01-05 | 江西师范大学 | Image query method based on feature classification |
CN112182262B (en) * | 2020-11-30 | 2021-03-19 | 江西师范大学 | Image query method based on feature classification |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103714122A (en) | Image retrieval method based on local block binary coding characteristics | |
Luvizon et al. | A video-based system for vehicle speed measurement in urban roadways | |
CN107392141B (en) | Airport extraction method based on significance detection and LSD (least squares distortion) line detection | |
CN101833664A (en) | Video image character detecting method based on sparse expression | |
CN102799867A (en) | Meter pointer angle identification method based on image processing | |
CN109919002B (en) | Yellow stop line identification method and device, computer equipment and storage medium | |
CN102509109B (en) | Method for distinguishing Thangka image from non-Thangka image | |
CN105260428A (en) | Picture processing method and apparatus | |
CN105957107A (en) | Pedestrian detecting and tracking method and device | |
CN104463138A (en) | Text positioning method and system based on visual structure attribute | |
CN104851089A (en) | Static scene foreground segmentation method and device based on three-dimensional light field | |
CN104598907A (en) | Stroke width figure based method for extracting Chinese character data from image | |
CN104463134A (en) | License plate detection method and system | |
CN105426816A (en) | Method and device of processing face images | |
Liang et al. | A new wavelet-Laplacian method for arbitrarily-oriented character segmentation in video text lines | |
CN103337073B (en) | A kind of two dimensional image threshold segmentation method based on three-dimensional entropy | |
CN102831428B (en) | Method for extracting quick response matrix code region in image | |
Bulugu | Algorithm for license plate localization and recognition for tanzania car plate numbers | |
CN102136060A (en) | Method for detecting population density | |
JP5027201B2 (en) | Telop character area detection method, telop character area detection device, and telop character area detection program | |
CN103136536A (en) | System and method for detecting target and method for exacting image features | |
CN111241911A (en) | Self-adaptive lane line detection method | |
CN105930813B (en) | A method of detection composes a piece of writing this under any natural scene | |
CN106951831A (en) | A kind of pedestrian detection tracking based on depth camera | |
CN104572680A (en) | Clothes retrieval method based on color matrix |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140409 |
|
RJ01 | Rejection of invention patent application after publication |