Depth Modeling With Spectral Selective Region Coding For Image Inpainting

摘要

Image inpainting, has an evolving approach for image quality enhancement and image visualization. In the process of image inpainting, pixels of similar area variants are considered in a tracing manner to achieve the objectives of unwanted image coefficient which are introduced due to detritions in image handing. To overcome this issue, images are processed in spatial domain, where, images are traced using 8-neighbor region growing method to achieve the objective of image enhancement However, in such approach, the pixel variations are observed in one variation plane. The variation with respect to successive pixel variants is not observed. To develop a new coding in considering with multiple domains, in this paper a new inpainting approach based on image depth coding is suggested.