CN107169968A - A kind of image block and processing method based on raster scanning principle - Google Patents
A kind of image block and processing method based on raster scanning principle Download PDFInfo
- Publication number
- CN107169968A CN107169968A CN201710177615.5A CN201710177615A CN107169968A CN 107169968 A CN107169968 A CN 107169968A CN 201710177615 A CN201710177615 A CN 201710177615A CN 107169968 A CN107169968 A CN 107169968A
- Authority
- CN
- China
- Prior art keywords
- image
- image block
- block
- low
- resolution
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 5
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000012545 processing Methods 0.000 claims abstract description 6
- 238000005192 partition Methods 0.000 claims description 9
- 239000000203 mixture Substances 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 5
- 238000002474 experimental method Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 3
- 230000003287 optical effect Effects 0.000 abstract 1
- 238000010894 electron beam technology Methods 0.000 description 2
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Television Systems (AREA)
Abstract
A kind of image block and processing method based on raster scanning principle, the present invention is inspired by raster scanning principle in optical system, by image by initial coordinate from left to right, several image blocks are divided into by tunable steps from top to bottom, and abstract image information is used for various image processing algorithms from these ready-portioned image blocks.Especially, this method is applied in super-resolution image reconstruction and image enhaucament by we, has been achieved with extraordinary effect.
Description
Technical field:
The invention belongs to digital image processing techniques field, and in particular to a kind of abstract image based on raster scanning principle
The image block method of middle information.
Background technology:
When raster scan display shows figure, electron beam is swept according to fixed scan line and defined scanning sequency
Retouch.Electron beam first since the fluorescent screen upper left corner, sweeps to the right a horizontal line, then promptly flyback is more on the lower side to the left side
Position, then Article 2 horizontal line is swept, fixed path and order are swept down like this, and to the last a horizontal line, that is, complete
The scanning of whole screen.The image block limited amount that traditional image block method is obtained, thus obtain image information also compared with
It is few, it is not well positioned to meet the requirement of each image processing algorithm.
The content of the invention:
The invention provides a kind of image block method of information in abstract image based on raster scanning principle, this method
More image blocks can be obtained, so as to obtain more image informations.Especially, this method is applied to oversubscription by us
In resolution Image Reconstruction and image enhaucament, extraordinary effect is had been achieved with.
A kind of image information abstracting method based on raster scanning, this method comprises the following steps:
Step 1, from (x, y) coordinate points sized images block is divided initially to sample image, step-length is s;
1≤x≤m, x=k*s+1, k ∈ N, m are the length of image block;
1≤y≤n, y=k*s+1, k ∈ N, n are the width of image block;
Step 2, from ready-portioned image block abstract image information be used for image processing algorithm.
Described image processing method is:In the experiment of super-resolution image reconstruction and image enhaucament, the piecemeal side is used
Method abstract image information, is comprised the following steps that:
1) piecemeal is carried out to high-low resolution training image respectively using above-mentioned method of partition, obtains one group of high-low resolution
Image block;
2) information of low-resolution image block is extracted as input vector, extracts the information conduct in high-definition picture block
Pixel tag, one group of vector of composition is right;
3) judge that these image blocks belong to high frequency spatial or low frequency space and use support vector regression using Log operators
It is vectorial right after the training optimization of machine (SVR) instrument, obtain two dictionaries under low-and high-frequency space;
4) extract in test low-resolution image and image block and obtain the input vector under low-and high-frequency space, utilize SVR
Instrument returns corresponding image after belonging to the label pixel of super-resolution image block and being returned;
Finally, 5) image post-process obtaining final super-resolution image;
More image blocks can be obtained using the method for partition, so as to extract more image information composition of vector pair.
The image block obtained using the method for partition, which is quantitatively had, to have great advantage, and can be extracted from image block more
Image information.Test result indicates that, compared with other representative method for reconstructing, the image obtained using the above method
3.1%-5.3% and 1.5%-8.1% is respectively increased in average specific bicubic interpolation algorithm in PSNR and SSIM indexs.
Brief description of the drawings
Fig. 1 gives the flow chart of the present invention.
Fig. 2 gives the implementation process of the present invention.
Embodiment
Embodiment 1
Reference picture 1,2 pairs of this method are described in detail.
The image block method of information, specifically includes following steps in abstract image based on raster scanning principle:
1st, to the images of a width 256*256 sizes respectively from (x, y) (x=1,2,3;Y=1,2,3) coordinate points start to be divided into
The image block of 3*3 sizes, step-length is set to 1.
2nd, 3*3=9 piecemeal, obtained image number of blocks respectively p are carried out altogether1=p2=p4=p5=85*85, p3=
p6=p7=p8=85*54, p9=84*84.
3rd, the quantity p=p of image block is finally given1+p2+…+p9。
4th, abstract image information is used for various image processing algorithms from ready-portioned image block.
We are in the experiment of super-resolution image reconstruction and image enhaucament, using the method for partition abstract image information,
Obvious effect is had been achieved for, specific experiment step is as follows:
1st, piecemeal is carried out to high-low resolution training image respectively using above-mentioned method of partition, obtains one group of high-low resolution
Image block.
2nd, the information of low-resolution image block is extracted as input vector, extracts the information conduct in high-definition picture block
Pixel tag, one group of vector of composition is right.
3rd, judge that these image blocks belong to high frequency spatial or low frequency space and use support vector regression using Log operators
It is vectorial right after the training optimization of machine (SVR) instrument, obtain two dictionaries under low-and high-frequency space.
4th, extract in test low-resolution image and image block and obtain the input vector under low-and high-frequency space, utilize SVR
Instrument returns corresponding image after belonging to the label pixel of super-resolution image block and being returned.
5th, it is last, final super-resolution image is obtained after handling image.
More image blocks can be obtained using the method for partition, so as to extract more image information composition of vector pair.
Test result indicates that, compared with other representative method for reconstructing, the image obtained using the above method in PSNR and
3.1%-5.3% and 1.5%-8.1% is respectively increased in average specific bicubic interpolation algorithm in SSIM indexs.
Claims (2)
1. a kind of image block method based on raster scanning principle, it is characterised in that this method comprises the following steps:
Step 1, from (x, y) coordinate points sized images block is divided initially to sample image, step-length is s;
1≤x≤m, x=k*s+1, k ∈ N, m are the length of image block;
1≤y≤n, y=k*s+1, k ∈ N, n are the width of image block;
Step 2, from ready-portioned image block abstract image information be used for image processing algorithm.
2. a kind of image processing method of the image block method based on raster scanning principle based on described in claim 1, its
It is characterised by:In the experiment of super-resolution image reconstruction and image enhaucament, using the method for partition abstract image information, specifically
Step is as follows:
1) piecemeal is carried out to high-low resolution training image respectively using above-mentioned method of partition, obtains one group of high-low resolution image
Block;
2) information of low-resolution image block is extracted as input vector, and the information extracted in high-definition picture block is used as pixel
Label, one group of vector of composition is right;
3) judge that these image blocks belong to high frequency spatial or low frequency space and use support vector regression using Log operators
(SVR) it is vectorial right after instrument training optimization, obtain two dictionaries under low-and high-frequency space;
4) extract in test low-resolution image and image block and obtain the input vector under low-and high-frequency space, utilize SVR instruments
Return it is corresponding belong to the label pixel of super-resolution image block and returned after image;
Finally, 5) image post-process obtaining final super-resolution image;
More image blocks can be obtained using the method for partition, so as to extract more image information composition of vector pair.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710177615.5A CN107169968A (en) | 2017-03-23 | 2017-03-23 | A kind of image block and processing method based on raster scanning principle |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710177615.5A CN107169968A (en) | 2017-03-23 | 2017-03-23 | A kind of image block and processing method based on raster scanning principle |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107169968A true CN107169968A (en) | 2017-09-15 |
Family
ID=59848834
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710177615.5A Pending CN107169968A (en) | 2017-03-23 | 2017-03-23 | A kind of image block and processing method based on raster scanning principle |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107169968A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103116880A (en) * | 2013-01-16 | 2013-05-22 | 杭州电子科技大学 | Image super resolution rebuilding method based on sparse representation and various residual |
CN103440675A (en) * | 2013-07-30 | 2013-12-11 | 湖北工业大学 | Overall situation reconstitution optimization model construction method for image block compressed sensing |
CN104574336A (en) * | 2015-01-19 | 2015-04-29 | 上海交通大学 | Super-resolution image reconstruction system based on self-adaptation submodel dictionary choice |
CN105430413A (en) * | 2015-11-17 | 2016-03-23 | 复旦大学 | Four-block hardware scanning method applicable for integer motion estimation in HEVC (High Efficiency Video Coding) standard |
-
2017
- 2017-03-23 CN CN201710177615.5A patent/CN107169968A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103116880A (en) * | 2013-01-16 | 2013-05-22 | 杭州电子科技大学 | Image super resolution rebuilding method based on sparse representation and various residual |
CN103440675A (en) * | 2013-07-30 | 2013-12-11 | 湖北工业大学 | Overall situation reconstitution optimization model construction method for image block compressed sensing |
CN104574336A (en) * | 2015-01-19 | 2015-04-29 | 上海交通大学 | Super-resolution image reconstruction system based on self-adaptation submodel dictionary choice |
CN105430413A (en) * | 2015-11-17 | 2016-03-23 | 复旦大学 | Four-block hardware scanning method applicable for integer motion estimation in HEVC (High Efficiency Video Coding) standard |
Non-Patent Citations (1)
Title |
---|
袁其平等: "《用支持向量回归法实现单帧图像超分辨率重建》", 《光学精密工程》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
DE102010046259A1 (en) | Image frequency conversion using motion estimation and compensation | |
CN108154474A (en) | A kind of super-resolution image reconstruction method, device, medium and equipment | |
RU2596062C1 (en) | Method for correction of eye image using machine learning and method of machine learning | |
CN106485656A (en) | A kind of method of image super-resolution reconstruct | |
US20210312598A1 (en) | Display apparatus and image processing method thereof | |
US20170206633A1 (en) | Method and apparatus for up-scaling an image | |
KR20120024799A (en) | Apparatus and method for identifying the creator of a work of art | |
CN106254888A (en) | A kind of picture coding and coding/decoding method, image processing equipment | |
US7327904B2 (en) | Pattern classification and filter design for increasing image resolution | |
CN112561913A (en) | Method and device for generating mura defect sample data of display panel | |
US11769195B2 (en) | Systems and methods for visualizing wall coverings in an image of a scene | |
CN104036498A (en) | Fast evaluation method of OCT image quality based on layer by layer classification | |
CN102769732B (en) | A kind of conversion method for realizing video field | |
CN106254878A (en) | A kind of picture coding and coding/decoding method, image processing equipment | |
CN107220934A (en) | Image rebuilding method and device | |
Bradley et al. | Virtual microscopy with extended depth of field | |
CN112954355B (en) | Image frame processing method and device | |
CN105992003A (en) | Method and device for image compression of palette color number based on order or frequency | |
US20120087582A1 (en) | Method and system for resizing an image | |
CN107169968A (en) | A kind of image block and processing method based on raster scanning principle | |
CN108282664A (en) | Image processing method, device, system and computer readable storage medium | |
CN107590194A (en) | The cutting loading display method of super large image | |
CN106327428A (en) | Image super-resolution method and system based on transfer learning | |
US11770496B2 (en) | Systems and methods for visualizing surface coverings in an image of a scene | |
CN106303534A (en) | Multiple index string and pixel string merge method for compressing image and the device of copy mode |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170915 |