Unlock instant, AI-driven research and patent intelligence for your innovation.

Image deconvolution using color priors

A deconvolution and image technology, applied in the field of image deconvolution and equipment using color prior, can solve the problem of ineffective arbitrary images

Active Publication Date: 2013-09-18
MICROSOFT TECH LICENSING LLC
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Previous deconvolution methods for deblurring are often limited to specific applications, are often ineffective for arbitrary images, and sometimes generate unwanted artifacts such as ringing

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
  • Image deconvolution using color priors
  • Image deconvolution using color priors
  • Image deconvolution using color priors

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0014] overview

[0015] The following description begins with some discussion of deconvolution theory and models for deblurring. The image prior is described next, followed by a discussion of gradient and color priors. This is followed by a discussion of color models and how to use them to find which colors to use as color priors.

[0016] The technical details described in this paper can be better understood in light of the following general observations on photographic images and resulting image models. Globally, most images have a relatively limited set of different colors. Furthermore, most small neighborhoods or localities in an image can be described by an even smaller set of colors, often even as few as two colors will suffice. A deblurred image can be modeled as a linear combination of two colors per pixel (ie, each pixel is a linear combination of two colors that varies from pixel to pixel). In other words, an image can be considered to be blended pixel by pixel ...

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

Described are techniques for image deconvolution to deblur an image given a blur kernel. Localized color statistics derived from the image to be deblurred serve as a prior constraint during deconvolution. A pixel's color is formulated as a linear combination of the two most prevalent colors within a neighborhood of the pixel. This may be repeated for many or all pixels in an image. The linear combinations of the pixels serve as a two-color prior for deconvolving the blurred image. The two-color prior is responsive to the content of the image and it may decouple edge sharpness from edge strength.

Description

Background technique [0001] A common problem in photography is image blur, which can be caused by a combination of camera shake during long exposure times, subject movement, use of large apertures in low light settings, or limited camera resolution. Regardless of the cause, image blur is generally undesirable. [0002] With the advent of digital photography, it became possible to reduce or correct blur in images. figure 1 A blurred image is shown that is subjected to a deblurring process to produce a deblurred image 102 . Various methods have been used to find a deblurred image, such as image 102 . Some approaches have attempted to modify how images are captured. Some methods have used and added information from multiple images to reduce blur. An upsampling algorithm is used to reduce blur from the limited camera resolution. A blur kernel determined from a single image is also used. Non-blind deconvolution has also been explored with limited success. [0003] Non-blind ...

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): G06T5/00
CPCG06T5/001G06T5/003G06T2207/10024G06T5/73
Inventor C·L·泽特尼克R·泽里斯基N·乔希
Owner MICROSOFT TECH LICENSING LLC