Depth Estimation From Stereo Image Based on Texture And Sparsity

Abstract

This paper presents a stereo image processing
techniques for the detection of a depth information from images.
Inspite of a remarkable growth in the last few years, the
availability of 3D content is still diminutive by that of its 2D
portion. To solve this problem, many 2D-to-3D image methods
have been proposed. Here a stereo image processing technique is
actually a 3D modeling technique. We analyze two types of
methods. The first is a texture mapping. For a detail like surface
texture and a color to extract from computer generated graphic
model or 3D model, is a technique called as texture mapping .
The second method is based on sparsity to obtain depth
information of stereo image. Finally comparison of both
methods are elaborated with parameters like PSNR and MSE.
Hence at the end it shows a better algorithm for stereo image.