AN EXTENDED EXEMPLAR BASED IMAGE INPAINTING TECHNIQUE FOR EFFECTIVE REGION FILLING USING TEXTURE PRIMITIVE

作者：K SANGEETHA; Dr P SENGOTTUVELAN

来源：International Journal of Engineering Science and Technology, 2012.

• Extended Exemplar-based Image Inpainting
• Mean Squared Error (MSE)
• Peak Signal-to-Noise Ratio (PSNR)
• Texture Primitive
• Image Inpainting.

In the field of multimedia signal processing, image inpainting is the technique commonly exploited for automatic filling-in the missing or unclear regions of an image without affecting the image quality. The region filling of the missing information is a exceedingly significant area in image processing, with large number of applications including image coding and wireless image transmission (e.g., recovering lost blocks), special effects (e.g., removal of objects) and image restoration (e.g., scratch removal). Inpainting techniques are also exploited to restore or enhance historical images which might have lost information through damage or scratch.Eliminating image interference or unnecessary portions of an image is another area inpainting. Exemplar-based inpainting is a technique which works well in cases of normal textures, where the lost information can be refilled by appropriate patches from the other areas of the same image. In this paper, the exemplar-based imageinpainting technique is extended by incorporating texture primitives for effective region filling. The main objective is to construct an image that appears to be %26apos;normal%26apos;, means that the viewer should not be able to find that the image has been inpainted. The performance is evaluated in terms of Mean Squared Error (MSE) and Peak Signal-to-Noise Ratio (PSNR).

• 出版日期2012