利用边缘像素深度特性的深度估计方法

Depth characteristic of the edge pixels based depth estimation method

为了提高物体边缘处像素的深度估计准确性,改善任意视点电视系统中合成虚拟视图质量,提出一种新颖的深度估计方法.该方法首先充分利用物体边缘像素的深度特性,将视图中的像素划分为3类,然后为每类像素提出相应的匹配代价函数的平滑项.与现有的基于边缘辅助信息的半自动深度估计方法相比,本方法有效提高了物体边缘处像素的深度估计准确性,在任意视点电视(FTV)系统接收端合成的虚拟视图客观质量PSNR平均提高0.32dB.

To improve the depth estimation precision at the object edges, a novel depth characteristic of the edge pixels based depth estimation method（DCE-SADE） is proposed . In the proposed method, based on the depth characteristics of the object edges, the pixels in the image are firstly divided into three categories, and then the corresponding smoothing terms of the matching cost function are derived. Experimental results have shown that compared with the existing edge information based semi-automatic depth estimation method（E-SADE） in the free- viewpoint television （FFV） system, the proposed method can greatly improve the depth estimation precision at the object edges, and that the PSNR of the synthesized virtual view based on the proposed method can be improved to a certain extent with an average of 0. 32 dB in the receiver of the FTV system.