Unsupervised texture segmentation using Gabor filters
Pattern Recognition
A Variable Window Approach to Early Vision
IEEE Transactions on Pattern Analysis and Machine Intelligence
Artificial Vision for Mobile Robots: Stereo Vision and Multisensory Perception
Artificial Vision for Mobile Robots: Stereo Vision and Multisensory Perception
A Stereo Matching Algorithm with an Adaptive Window: Theory and Experiment
IEEE Transactions on Pattern Analysis and Machine Intelligence
Stereo Matching with Implicit Detection of Occlusions
ECCV '98 Proceedings of the 5th European Conference on Computer Vision-Volume II - Volume II
| Hi-index | 0.00 |
Area-based stereo matching algorithms are based on the support of a surrounding area to establish a point-to-point correspondence in two images. Two important problems arise in this context: How to obtain subpixel information and how to choose the optimal surrounding area. In this paper we present a non-iterative two-step algorithm for subpixel accurate stereo matching by using an adaptive window. In contrast to existing algorithms the window is not restricted to a rectangle but can be of any general shape. Starting from an initial sparse disparity estimate, the first step is to find the general shape of the window. This is performed by estimating the local disparity of each pixel in a box of maximum size using a bank of Gabor filters, and by applying a consistency constraint. In the second step the projective distortion is computed using the masked window. The performed experiments show the accurate and robust behavior of the proposed algorithm.