Translator Disclaimer
13 October 2008 Identification of blur support size in blind image restoration with moderate/intense noise
Author Affiliations +
This paper makes improvements on the algorithm of identifying blur support size proposed by Li Chen in blind image restoration to accommodate to the moderate/intense noise circumstance. The prior method mainly constructs a whiten filter referring to the image characteristics based on ARMA (Autoregressive Moving Average) model, and calculates the correlation of whiten filtered image with different shifts, thus gets the estimated blur size equal to the shifts at the minimum of correlation. However, this method is difficult to accurately identify blur size even with moderate additive noise. Some procedures are taken before calculating the correlation to ameliorate the estimation accuracy, including regarding edge neighborhoods as valid regions for calculating correlation in the whiten filtered image and implementing filtering to further reduce noise interference. Experimental results represent that the correlation attains its minimum when the shift distance meets certain relationship with actual blur size, thus show the improvements are valid for the case of moderate/intense contamination.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Chunxiao Zhang, Yan Zhao, and Dong Xu "Identification of blur support size in blind image restoration with moderate/intense noise", Proc. SPIE 7129, Seventh International Symposium on Instrumentation and Control Technology: Optoelectronic Technology and Instruments, Control Theory and Automation, and Space Exploration, 71290S (13 October 2008);

Back to Top