Abstract | ||
---|---|---|
This study proposes a data separation algorithm with computationally efficient strategies for non-blind deconvolution in the frequency domain. First, the blurred image is separated into a couple of basis and corresponding coefficients. Then the traditional non-blind deconvolution method in the frequency domain is employed in the pre-processing step and the deblurred bases are saved. For the new blurred image with the same blur kernel, the iterative optimization is converted into linear addition and multiplication operations. Results of qualitative and quantitative evaluations demonstrate the efficiency and effectiveness of the proposed approach. This method is meaningful for the design of a real-time image deconvolution method. Thus far, this method is only suitable in the frequency domain. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1145/3015166.3015169 | ICSPS |
Field | DocType | Citations |
Kernel (linear algebra),Frequency domain,Computer vision,Deblurring,Blind deconvolution,Matrix decomposition,Deconvolution,Wiener deconvolution,Algorithm,Artificial intelligence,Image restoration,Mathematics | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Weili Li | 1 | 4 | 2.41 |
Yu Liu | 2 | 7 | 4.48 |
Xiaoqing Yin | 3 | 0 | 2.03 |
Maojun Zhang | 4 | 314 | 48.74 |