Development and Analysis of Enhanced Image Inpainting Approach

Inpainting is a restoration technique that involves filling in damaged, degraded, or missing areas of artwork to create a full image. Oil or acrylic paints, biochemical photography prints, sculptors, and digital photos and video are all examples of physical and digital art mediums that can be used in this process. We have developed a model which in paint a corrupted image using three different sparse representation based approaches. We have used K SVD Singular Value Decomposition, ORTHOGONAL MATCHING PURSUIT OMP , and Delaunay Triangulation based Interpolation. It has also become apparent from the results that inpainting on natural images appears decent when not requiring too big patch size. As the patch sizes increase, to be able to cover large masked areas, the reconstruction will be smoother. There might be useful at times to have an algorithm like this that could be used for smoothing and inpainting simultaneously. However, if the contrast of the image is to be unharmed some other method should be considered. It is therefore concluded that even though sparse reconstructive methods appear impressive at first glance they are lacking when it comes to using large image patches, which is required when it comes to inpainting phase maps. Yogita | Shalini Bhadola | Kirti Bhatia | Rohini Sharma "Development and Analysis of Enhanced Image Inpainting Approach" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-6 | Issue-4 , June 2022, URL: https://www.ijtsrd.com/papers/ijtsrd50366.pdf Paper URL: https://www.ijtsrd.com/computer-science/cognitive-science/50366/development-and-analysis-of-enhanced-image-inpainting-approach/yogita