摘要: This paper describes a robust, pixel-based adaptive algorithm to track imagefeatures, both spatially and temporally, over a sequence of monocular images. The algorithm assumes no a priori knowledge about the features to be tracked... 展开 This paper describes a robust, pixel-based adaptive algorithm to track imagefeatures, both spatially and temporally, over a sequence of monocular images. The algorithm assumes no a priori knowledge about the features to be tracked, or the motion of the camera or the objects, in the 3-D scene. The features to be tracked are selected by the algorithm and these correspond to the peaks of a 'matching surface' constructed from the first image of the sequence to be analyzed. Any kind of motion can be tolerated keeping in mind the pixels-per-frame motion limitations. In our experiments, we have used the algorithm for tracking up to 3 pixels motion between frames with absolutely no subpixel calculations being necessary. Taking into account constraints of temporal continuity, the algorithm uses simple and efficient predictive tracking over multiple frames. The algorithm accepts a slow, continuous change of brightness D.C. level in the pixels of the feature. Trajectories of features on multiple objects can be computed. As an application of the algorithm we show how the algorithm can be used to find the Focus of Expansion (FOE) using a sequence of real images. 收起