Li Yunpeng, Ge Baozhen, Tian Qingguo, Lu Qieni, Quan Jianing, Chen Qibo, Chen Lei
Opt Express. 2021 Oct 11;29(21):33874-33889. doi: 10.1364/OE.440241.
It is challenging to realize stereo matching in dynamic stereo vision systems. We present an epipolar guided optical flow network (EGOF-Net) for unrectified stereo matching by estimating robust epipolar geometry with a deep cross-checking-based fundamental matrix estimation method (DCCM) and then surpassing false matches with a 4D epipolar modulator (4D-EM) module. On synthetic and real-scene datasets, our network outperforms the state-of-the-art methods by a substantial margin. Also, we test the network in an existing dynamic stereo system and successfully reconstruct the 3D point clouds. The technique can simplify the stereo vision pipeline by ticking out rectification operations. Moreover, it suggests a new opportunity for combining heuristic algorithms with neural networks. The code is available on https://github.com/psyrocloud/EGOF-Net.
在动态立体视觉系统中实现立体匹配具有挑战性。我们提出了一种极线引导光流网络（EGOF-Net），用于通过基于深度交叉检查的基本矩阵估计方法（DCCM）估计稳健的极线几何，然后用4D极线调制器（4D-EM）模块超越错误匹配，从而实现未校正的立体匹配。在合成数据集和真实场景数据集上，我们的网络大幅超越了现有最先进的方法。此外，我们在现有的动态立体系统中测试了该网络，并成功重建了三维点云。该技术可以通过剔除校正操作来简化立体视觉流程。此外，它为启发式算法与神经网络的结合提供了新的机会。代码可在https://github.com/psyrocloud/EGOF-Net上获取。
