一种视差法与变焦法融合的深度估计方法  

A Fused Estimation Method of Depth Based on Depth from Stereo and Depth from Focus

作　　者：肖照林[1] 杨恒[1] 周果清[1] 王庆[1] 

机构地区：[1]西北工业大学计算机学院,陕西西安710072

出　　处：《西北工业大学学报》2013年第2期311-316,共6页Journal of Northwestern Polytechnical University

基　　金：国家自然科学基金(61103060;61272287);教育部博士点基金(20116102110031)资助

摘　　要：视差法与变焦法是两种基于图像的三维重建方法,由于它们在所需数据集与深度计算方法上的差异,故具有不同的优势和应用场合。通过利用摄像机阵列,在获取多视点数据集的基础上,提出了一种视差法与变焦法融合的深度估计方法。先利用视差法选取候选深度集,然后利用视差法与变焦法结果的带权线性混合函数选取最终深度,从而解决了视差法与变焦法数据集的获取及单一方法对深度不连续区域鲁棒性较低的问题。实验结果表明该方法的可行性与优越性。Depth from stereo and depth from focus are two classical image-based 3 D reconstruction methods, which have distinct advantages in different applications respectively. In this paper, we propose a fused depth recovery method based on depth from stereo and depth from focus. First, we choose some depth candidates using depth from stereo method. Then, a weighted linear cost function is employed to determine the final depth. As we know, depth from stereo and depth from focus need different kinds of input images, which can not he acquired at the same time by using traditional imaging system. We have preliminary tackled this issue by using a camera array system and synthetic aperture photography. Experiment results have verified the validity and advantages of proposed method.

关 键 词：深度估计 视差法 变焦法 摄像机阵列 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]