An enhanced method for object removal using exemplar-based image inpainting

Image inpainting, also known as disocclusion or image completion is a method of reconstruction of lost or damaged region of an image or video. The art of inpainting is as ancient as art itself. It is done by professional artists in museums. In this paper, an enhanced method is proposed for the object removal using exemplar-based image inpainting. In this method, the image is normalized column wise and then the priorities of each pixel from the patch are calculated. The pixel is then replaced with the pixel which has maximum estimated priority. The replacement is continued until all the pixels of the desired object are removed. The results of numerical analysis show that our method outperforms than existing method. The accuracy check analysis gives up to 60% improvement.