Depth Enhancement and Optimization for Stereo Images Based on JNDD

Abstract

With rapid progress of three-dimensional television and 3D film industry, stereoscopic technology has become an increasingly important aspect of image technology in recent years. Therefore, in this paper, processing of stereo images is mainly focused on. Based on this new Just Noticeable Depth Difference (JNDD) model, a Modified Local Depth Enhancement Algorithm is proposed to enhance the depth of stereo images more effectively. And then we design a Local Depth Enhancement Algorithm based on Zero Parallax Plane (ZPP) depth optimization. While Modified Local Depth Enhancement Algorithm (MLD EA) enhancing effectively the depth perception compared with the traditional algorithm, ZPP depth optimization based Local Depth Enhancement Algorithm (ZLDEA) achieves a higher comfort score in subjective evaluation experiment.