Exemplar Based Image Inpainting using Pattern Extraction Approach & Restoration using Improved JSM Technique

Abstract: Image inpainting issue is a important amongst the most crucial picture recreation issues. It is the strategy of adjusting a picture in an imperceptible structure to a spectator not acquainted with the first image and is as outdated as craftsmanship itself. It likewise suggest to the act of the craftsmen of reestablishing works of art. Analysts chipping away at various applications have embraced distinctive names: picture addition, disocclusion, picture substitution, and blunder equipment however each of them conveys its own individual qualities. The proposed technique inpainted the image using Pattern extraction technique and Remove noise perform de-blurring with the help of JSM (Jonit Statistical Modeling) Technique. Proposed system works in four modules which are Image Processing, Deblurring Image, Text Removal, Mixed Noise Removal. Evaluation of the proposed system is done on colored images using exemplar based image inpainting technique. The comparison between existing method and proposed method is based on three parameters MSE, PSNR and Time and it is shown that proposed method improves these parameter values.

Keywords: Image Inpainting; Noise Removal; Image Processing; Image Deblurring; Image Restroration.