Efficient Stereo Matching Leveraging Deep Local and Context Information
Efficient Stereo Matching Leveraging Deep Local and Context Information
复制标题
利用深层局部和上下文信息进行高效立体匹配
DOI：
10.1109/access.2017.2754318
复制
发表时间：
2017-09
期刊：
影响因子：
3.9
通讯作者：
Zhang Xiaolin
中科院分区：
文献类型：
--
作者：
Ye Xiaoqing;Li Jiamao;Wang Han;Huang Hexiao;Zhang Xiaolin
立体匹配是一个具有挑战性的问题，对于弱纹理，不连续性，照明差异和遮挡。因此，本文提出了一个深度学习框架，重点关注典型立体方法的第一和最后阶段：匹配成本计算和视差细化。对于匹配成本计算，利用两种基于补丁的网络架构来允许速度和准确性之间的权衡，这两种架构都利用多尺寸和多层池化单元，没有步幅来学习跨尺度特征表示。对于视差细化，与传统的手工细化算法不同，我们在离群点检测之前将初始最优和次优视差图合并。此外，不同的基础学习者被鼓励专注于特定的替代任务，对应于平滑的区域和细节。在不同数据集上的实验结果表明了该方法的有效性，能够获得亚像素精度，并在很大程度上恢复遮挡。具体来说，我们的准确框架在非遮挡和遮挡区域都达到了近峰值的准确度，我们的快速框架在Middlebury基准测试中与快速算法相比具有竞争力的性能。
Stereo matching is a challenging problem with respect to weak texture, discontinuities, illumination difference and occlusions. Therefore, a deep learning framework is presented in this paper, which focuses on the first and last stage of typical stereo methods: the matching cost computation and the disparity refinement. For matching cost computation, two patch-based network architectures are exploited to allow the trade-off between speed and accuracy, both of which leverage multi-size and multi-layer pooling unit with no strides to learn cross-scale feature representations. For disparity refinement, unlike traditional handcrafted refinement algorithms, we incorporate the initial optimal and sub-optimal disparity maps before outlier detection. Furthermore, diverse base learners are encouraged to focus on specific replacement tasks, corresponding to the smooth regions and details. Experiments on different datasets demonstrate the effectiveness of our approach, which is able to obtain sub-pixel accuracy and restore occlusions to a great extent. Specifically, our accurate framework attains near-peak accuracy both in non-occluded and occluded region and our fast framework achieves competitive performance against the fast algorithms on Middlebury benchmark.
登录
查看更多内容
DOI：
10.1109/3dimpvt.2011.24
发表时间：
2011-05
期刊：
2011 International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission
影响因子：
--
作者：
Xun Sun;Xing Mei;Shaohui Jiao;Mingcai Zhou;Haitao Wang
通讯作者：
Xun Sun;Xing Mei;Shaohui Jiao;Mingcai Zhou;Haitao Wang
DOI：
10.1007/978-3-642-15549-9_1
发表时间：
2010-09
影响因子：
23.6
作者：
Kaiming He;Jian Sun-;Xiaoou Tang
通讯作者：
Kaiming He;Jian Sun-;Xiaoou Tang
DOI：
10.1109/cvpr.2006.292
发表时间：
2006-06
期刊：
2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)
影响因子：
--
作者：
Qingxiong Yang;Liang Wang;Ruigang Yang;Henrik Stewénius;D. Nistér
通讯作者：
Qingxiong Yang;Liang Wang;Ruigang Yang;Henrik Stewénius;D. Nistér
影响因子：
6
作者：
J. Bedő;Cheng Soon Ong
通讯作者：
J. Bedő;Cheng Soon Ong
影响因子：
19.5
作者：
C. Pal;Jerod J. Weinman;Lam C. Tran;D. Scharstein
通讯作者：
C. Pal;Jerod J. Weinman;Lam C. Tran;D. Scharstein