Computing the Stereo Matching Cost with a Convolutional Neural Network
This work addresses depth extraction for computer vision applications, representing an incremental improvement in stereo matching methods.
The paper tackled stereo depth estimation by training a convolutional neural network to compute matching costs, achieving an error rate of 2.61% on the KITTI dataset and top performance at the time.
We present a method for extracting depth information from a rectified image pair. We train a convolutional neural network to predict how well two image patches match and use it to compute the stereo matching cost. The cost is refined by cross-based cost aggregation and semiglobal matching, followed by a left-right consistency check to eliminate errors in the occluded regions. Our stereo method achieves an error rate of 2.61 % on the KITTI stereo dataset and is currently (August 2014) the top performing method on this dataset.