当前位置: 首页 >  论文展示 >  正文

HPM-TDP: An Efficient Hierarchical PatchMatch Depth Estimation Approach using Tree Dynamic Programming

虞敏 2019-07-02 浏览

Mao Tian, Bisheng Yang*, Chi Chen*, Ronggang Huang, and Liang Huo. “HPM-TDP: An Efficient Hierarchical PatchMatch Depth Estimation Approach using Tree Dynamic Programming ”, ISPRS Journal of Photogrammetry and Remote Sensing (Accepted), Jun. 2019.

该论文提出了一种高效、鲁棒的层次PatchMatch深度计算方法,克服了传统立体匹配方法平行平面(fronto-parallel)影响,提高了传统匹配方法的精度和效率,可广泛应用于城市场景三维重建、自动驾驶及机器人导航。

Fig. 1 The workflow of HPM-TDP.

Fig. 2. Comparison of the disparity maps of the KITTI 2015 dataset.