Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for reestablishment of single frame image quick super-resolution based on nucleus regression

A technology of super-resolution reconstruction and kernel regression, applied in image enhancement, image data processing, 2D image generation, etc., can solve the problems of long time consumption and large amount of calculation, and achieve the goal of improving processing speed, saving processing speed, highlighting Effects of nonlinear processing performance

Inactive Publication Date: 2009-08-26
江苏美梵生物科技有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a fast super-resolution reconstruction method for a single-frame image based on kernel regression, to overcome the defects of the existing super-resolution reconstruction method for a single-frame image with kernel regression, which has a huge amount of calculation and takes a long time

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for reestablishment of single frame image quick super-resolution based on nucleus regression
  • Method for reestablishment of single frame image quick super-resolution based on nucleus regression
  • Method for reestablishment of single frame image quick super-resolution based on nucleus regression

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0007] Specific implementation mode one: the following combination figure 1 , figure 2 , Figure 6 and Figure 7 This embodiment will be specifically described. This embodiment includes the following steps: 1. Map the pixel points on the low-resolution image to the high-resolution grid, and make the above-mentioned pixel points be located on the grid intersection of the high-resolution grid; 2. In the high-resolution grid, the grid intersections other than the grid intersections occupied by low-resolution image pixels are eliminated, as the pixel points to be evaluated, and the pixel points to be evaluated are further divided into two categories according to the spatial position relationship. The pixels to be evaluated are the remaining pixels to be evaluated after removing the points on the connection line between the pixels of the low-resolution image in the intersection points of the high-resolution grid; the second type of pixels to be evaluated are The pixels to be ...

specific Embodiment approach 2

[0036] Specific implementation mode two: the following combination Figure 5 and Figure 6 This embodiment will be specifically described. The difference between this embodiment and Embodiment 1 is that in step 3, the square neighborhood pixel set of the first type of pixels to be evaluated is determined as follows: a. Use a square local window and determine the size of the local window as n×n, n×n is the number of pixels in the low-resolution image in the local window, n is equal to 4 or 8, and choosing an even window of 4x4 can achieve a better reconstruction result, and the calculation speed is fast; b. Make the pixels to be evaluated Located in the center of the local window; c. All the pixels of the low-resolution image in the local window form a local neighborhood pixel set. by Figure 5 For example, when the pixel to be evaluated is X1, under the 4x4 window, the local neighborhood pixel set should be {A1, A2, A3, A4, B1, B2, B3, B4, C1, C2, C3, C4, D1 , D2, D3, D4};...

specific Embodiment approach 3

[0037] Specific implementation mode three: the following combination Figure 7 This embodiment will be specifically described. The difference between this embodiment and Embodiment 1 is that in step 4, the diamond-shaped neighborhood pixel set of the second type of pixel points to be evaluated is determined as follows: a. Use a diamond-shaped local window and determine the size of the local window as m×m, m is equal to 4 or 8, and m is the sum of the number of the first type of pixels to be evaluated and the number of low-resolution image pixels contained in one side of the rhombus. by Figure 6 For example, when the pixel point to be evaluated is X2, under the 4x4 window, the local neighborhood pixel set is the intersection point of the thick dotted line in the figure, that is, the hollow circle point and the hollow triangle point; it should be noted that the hollow triangle point here belongs to The first type of points whose estimated values ​​have been obtained previousl...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A fast super-resolution reconstruction method for a single-frame image based on kernel regression, the invention relates to a method for image super-resolution reconstruction. It overcomes the shortcomings of the existing super-resolution reconstruction method of kernel regression single-frame image, which is computationally intensive and time-consuming. It includes the following steps: map the pixels on the low-resolution image to the high-resolution grid; determine the pixels to be evaluated and divide them into two categories; determine the square neighbors of each first type of pixels to be evaluated Domain pixel set, the pixel value of each point in the set is substituted into the kernel regression equation to calculate the pixel value; the diamond-shaped neighborhood pixel set of the second type of pixels to be evaluated is determined, and the set is substituted into the kernel regression equation to calculate the pixel value; when all the pixels to be estimated After the value pixels are assigned, the image is output. The present invention introduces two-dimensional nonlinear kernel regression to estimate interpolation points, uses local neighborhood processing instead of whole image processing, and adopts an instant update strategy, thereby realizing super-resolution reconstruction of a single frame image.

Description

technical field [0001] The invention relates to a method for image super-resolution reconstruction. Background technique [0002] Spatial resolution is a measure of the imaging system's ability to distinguish image details, and it is also an indicator of the subtlety of objects in the image. However, in the process of image acquisition, many factors will lead to the decline or degradation of image quality. An effective way to solve this problem is super-resolution image reconstruction. Super-resolution image reconstruction is an image processing method that has emerged in recent years. This method obtains high-resolution images by estimating and integrating image information. It is an economical and easy-to-implement image reconstruction and resolution enhancement method. . The image super-resolution reconstruction technology uses the two-dimensional sampling values ​​of the degraded low-resolution image to undergo a series of two-dimensional operations to improve its res...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06T5/00G06T11/00
Inventor 谷延锋张晔
Owner 江苏美梵生物科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com