Author : Mahroosh Banday and Anuj Sharma

In the field of multimedia processing, the image inpainting technology is a hot spot. Inpainting is the process of filling in the missing/unclear regions of an image utilizing spatial information from its neighbouring region so as to preserve its overall continuity in a way that the changes made are not easily detected by an ordinary observer. Digital image inpainting is a new and young research area in image processing with numerous applications including wireless image transmission (recovering lost blocks), red eye removal, multimedia editing (special effects in movies) and image restoration. Eliminating unnecessary portions of an image (removal of objects) is another exceedingly significant area of digital inpainting. Exemplar-based image inpainting is the technique that works well in this area, where the removed/lost information can be filled in by selecting appropriate patches from the neighbouring areas of the same image. This paper presents a brief survey of Exemplar-based inpainting techniques and compares and analysis some recent Exemplar-based inpainting methods. The performance of these methods is compared using both qualitative approach as well as quantitatively in terms of processing speed, Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE).

Keywords: Inpainting, Exemplar-based, Texture Primitive, PSNR, MSE

Article published in International Journal of Current Engineering and Technology, Vol.4, No.5 (Oct-2014)