Title :
Image denoising based on steepest descent OMP and K-SVD
Author :
Xiangyu Deng;Zengli Liu
Author_Institution :
Faculty of Information Engineering and Automation, KunMing University of Science and Technology, Kunming, China
Abstract :
Noise suppression is one of the key problems in image processing. In recent years, sparse representation theory is applied in image denoising successfully. The primary idea is to denoise an image via over-complete dictionary trained by K-SVD algorithm based on OMP (Orthogonal Matching Pursuit) algorithm. This method receives good performance on the quality of image denoising but slow computation speed because of high computational complexity. In order to speed up the computation while keeping image quality. This paper discusses a denoising method via the adaptive over-complete dictionary trained from noisy image using improved K-SVD algorithm and the steepest descent OMP algorithm. In this work, we replace OMP with the steepest descent OMP. Simulation results show this method leads to a better balance between denoising quality and the computation speed, and can improve performance than other methods. The PSNR values are used to measure the denoising quality, and it has been proven the PSNR values can be increased by our method meanwhile the running time can also be reduced to some extent.
Keywords :
"Dictionaries","Matching pursuit algorithms","Image denoising","Noise reduction","Noise measurement","Transforms","Image reconstruction"
Conference_Titel :
Signal Processing, Communications and Computing (ICSPCC), 2015 IEEE International Conference on
Print_ISBN :
978-1-4799-8918-8
DOI :
10.1109/ICSPCC.2015.7338821