We describe a method to solve the stereo correspondence using controlled head (or camera) movements. These movements, which can be due to eye rotation, head rotation, or head translation, essentially supply additional imageframes which can be used to constrain the stereo matching by supplying monocular cues. Because the movements are small, traditional methods of stereo with multiple frame will not work. We develop an alternative approach using a systematic analysis to define a probability distribution for the errors. Our matching strategy then matches the most probable points first (based on the monocular cues), thereby reducing the ambiguity for the remaining matches. We demonstrate this algorithm in detail for the cases of head and eye rotation and illustrate it with some examples.
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition
- Artificial Intelligence