A locally global approach to stereo correspondence

A novel approach to deal with the stereo correspondence problem induced by the implicit assumptions made by cost aggregation (CA) strategies is proposed. CA relies on the implicit assumption that disparity varies smoothly within neighboring points except at depth discontinuities and state-of-the-art CA strategies adapt their support to image content by classifying each pixel based on geometric and photometric constraints. Our proposal explicitly models this behavior from a different perspective, by gathering for each point, multiple assumptions that locally would be made by a hypothetical variable CA strategy. This framework enables to derive a function that locally captures the plausibility of the underlying geometric and photometric constraints independently enforced by supports of neighboring points. Experimental results confirm the effectiveness of our proposal.