Open Access Open Access  Restricted Access Subscription or Fee Access

Reconstruction of Images using Image Inpainting and Super Resolution

Manish Agate, M. P. Deshmukh, Gaurav Brahmbhatt, Anmol Nimbalkar, Sanket Sakhare, Akib Peerzade

Abstract


Image completion of large missing regions in a selected image is a challenging task. Old photographs have some part of the image distorted or some part of it is ruined by folds and creases. Hiding these flaws or covering such missing regions is a necessity sometimes and needs to be solved by using some advanced techniques. Inpainting is one such technique that helps to overcome these problems. Inpainting is a technique of modifying an image in an undetectable form. Using inpainting requires the user to select a part of the image to be removed and the algorithms used in it will restore the required part or patch. Image inpainting methods can be classified into two main categories viz. Diffusion based and Examplar based approaches. Diffusion based approach has varying methods and mathematical evaluations involved. Examplar based approach samples and copies best matching samples from the neighboring pixels of the image surrounding the selected patch. A recent approach in image inpainting is a combination of two methods namely examplar based approach and single image super resolution. The super resolving is done after the examplar based approach finishes its operations so that the patching, which is coarse, gets the finishing touch by super resolving it. If the area to be restored is large then the diffusion based approach introduces some blurring whereas examplar based approach can handle large areas effectively.

Keywords: Examplar based, inpainting, super resolution


Full Text: PDF

Refbacks

  • There are currently no refbacks.