DE-Mosaicing using Matrix Factorization Iterative Tunable Method
Shabana Tabassum1, SanjayKumar C. Gowre2
1Shabana Tabassum*, Associate Professor, Dept of E&CE, KBN College of Engineering, Kalaburagi, Karnataka, India.
2Dr. SanjayKumar C. Gowre, Associate Professor, Dept of E&CE, KBN College of Engineering, Kalaburagi, Karnataka, India.
Manuscript received on October 12, 2019. | Revised Manuscript received on 21 October, 2019. | Manuscript published on November 10, 2019. | PP: 473-480 | Volume-9 Issue-1, November 2019. | Retrieval Number: A4213119119/2019©BEIESP | DOI: 10.35940/ijitee.A4213.119119
Open Access | Ethics and 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: A color image is captured though the single image sensor and it is named as the mosaicked image and this is obtained through the CFA where the pixels are arranged such that any one of the color from the given color component is recorded at every pixel. De-mosaicingis absolute reverse of mosaicking, where the process is to reconstruct the full color image from the given incomplete color samples. In past several methods of de-mosaicing have been proposed however, they have many shortcomings such as computational complexity and high computational load, not matching the original images. Hence we have proposed a methodology named as MFIT i.e. Matrix factorization Iterative Tunable approach, the main aim of this methodology is to improvise the reconstruction quality. In order to achieve the better reconstruction quality we have used the MFIT algorithm at each iteration this helps in avoiding the image artifacts and it is achieved through the image block adjustment also it reduces the computational load. Moreover in order to evaluate the proposed algorithm we have compared it with nearly 12 algorithm based on the value of PSNR and SSIM, the theoretical results and comparative analysis shows that our algorithm excels compared to the other existing method of de-mosaicing.
Keywords: De-mosaicing, MFIT, Matrix Factorization, Iterative Tunable.
Scope of the Article: Signal and Image Processing