PSF Estimation with PSO and SURE LET Deconvolution for Blurred Image
Ritesh Pawar1, Maiytree Dutta2
1Ritesh Pawar, M.E Scholar, Department of Electronics & Communication Engineering, National Institute of Technical Teacher Training & Research, Chandigarh (UT), India.
2Dr. Maiytree Dutta, Professor & Head, Department of Electronics & Communication Engineering, National Institute of Technical Teacher Training & Research, Chandigarh (UT), India.
Manuscript received on 12 November 2016 | Revised Manuscript received on 24 November 2016 | Manuscript Published on 30 November 2016 | PP: 19-24 | Volume-6 Issue-6, November 2016 | Retrieval Number: F2391116616/16©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this research we proposed a technique for the Point spread function estimation in the form of the particle swarm optimization, here also use unbiased risk estimation for the MSE in filtered version with blur stein’s unbiased risk estimation in the form of the novel criterion to calculate only PSF from the blurred image which is unknown. This process of minimization of PSF is obtained by the wiener filtering. On the estimation of this blur kernel, non blind deconvolution is done with the SURE LET deconvolution algorithm. The best positions of the particles are calculated by the PSO. Here we use gaussian kernel for parametric form. In this research we found that position calculation from PSO gives the more accurate PSF parameter estimations, this may lead the high accuracy in restoration of degraded images which is as similar to the exact PSF, when whole result is performed with the help of the SURE LET deconvolution algorithm. From the result it is found that non blind deconvolution has highly accurate results in the form of the visually and computationally form.
Keywords: PSF Estimation, PSO, Exact Wiener Filtering, SURE LET, Blur SURE.
Scope of the Article: Image Security