IMAGE INPAINTING USING DISCRETE WAVELET TRANSFORMATION TECHNIQUE

Image inpainting problem is one of the most essential image restoration problems. It is the technique of modifying an image in an undetectable form to an observer not familiar with the original image and is as ancient as art itself. It also refers to the practice of the artists of restoring paintings. Researchers working on different applications have adopted different names: image interpolation, disocclusion, image replacement, and error concealment though each of them carries its own individual characteristics. ). Exiting system has a very large number of problems which include lower PSNR values, high MSE, large time and fixed region of image to inpaint etc. all these problems are required to be removed in the proposed work. In the proposed system, pattern extraction technique is used to find the best matching patch. In the proposed system image is inpainted on the basis of patch priority. Existing systems for image inpainting works only on basis of neibhoring pixels but the proposed system works on the basis of patch priprity of the neibhoring pixels. Proposed system follow the patteren on the basis of which it fills the targeted region of the image. In this paper Exempler based approach along with DWT is used to inpaint the targeted pixels.