Enhancement of Depth Map Using Texture and Depth Consistency

Presently, the procedure of we obtained depth map is based the stereo matching technique and stereo camera, but the depth map will be acquired based on the above method, the draft depth map is low resolution than texture image. it often produces regions with missing pixels. The missing pixel regions will not include any depth information for the depth image. This reason would lead the performance to degrade seriously in modern-day three-dimensional (3D) video applications. In this paper, we propose adaptive texture-similarity-based hole filling (ATSIIF) and rotating counsel depth enhancement (RCDE). The proposed system, which is used for the enhancement of depth maps, is achieved by suppressing the noise, filling holes and sharpening object edges simultaneously. Results of experiments, presented for still images, show that proposed technique increases the quality of depth, especially around the object boundary. Beside, a comparison of the performance of the ATSHF and RCDE with those of customary depth enhancement methods shows that the proposed system is efficient.