PSF Estimation using Sharp Edge Prediction

 
Neel Joshi Richard Szeliski David Kriegman
UCSD Microsoft Research UCSD


CVPR 2008


 

Sharp Edge Prediction: A blurry image (top left) and the 1D profile normal to an edge (top right, blue line). We predict a sharp edge (top right, dashed line) by propagating the max and min values along the edge profile. The algorithm uses predicted and observed values to solve for a PSF. Only observed pixels within a radius R are used. (bottom left) Predicted pixels are blue and valid observed pixels are green. (bottom right) The predicted values.

 

Abstract

 

Image blur is caused by a number of factors such as motion, defocus, capturing light over the non-zero area of the aperture and pixel, the presence of anti-aliasing filters on a camera sensor, and limited sensor resolution. We present an algorithm that estimates non-parametric, spatially-varying blur functions (i.e., point-spread functions or PSFs) at subpixel resolution from a single image. The method handles blur due to defocus, slight camera motion, and inherent aspects of the imaging system. The algorithm can be used to measure blur due to limited sensor resolution by estimating a sub-pixel, super-resolved PSF even for in-focus images. It operates by predicting the “sharp” version of a blurry input image and uses the two images to solve for a PSF. We handle the cases where the scene content is unknown and also where a known printed calibration target is placed in the scene. The method is completely automatic, fast, and produces accurate results.

 

Paper

Adobe Acrobat PDF (7.21 MB)

   

Copyright 2008 by Neel Joshi, UCSD, and Microsoft Research

Use of images, videos, and slides for non-commercial, academic, or news related purposes is allowed with proper attribution. Contact Neel Joshi for inquiries about other uses.