In this paper, we describe a local-neighborhood pixel-based adaptive a
lgorithm to track image features, both spatially and temporally, over
a sequence of monocular images. The algorithm assumes no a priori know
ledge about the image features to be tracked, or the relative motion b
etween the camera and the three dimensional(3D) objects. The features
to be tracked are selected by the algorithm and they correspond to the
peaks of a 'correlation surface' constructed from a local neighborhoo
d in the first image of the sequence to be analysed. Any kind of motio
n, i.e., 6 DOF (translation and rotation), can be tolerated keeping in
mind the pixels-per-frame motion limitations. No subpixel computation
s being necessary. Taking into account constraints of temporal continu
ity, the algorithm uses simple and efficient predictive tracking over
multiple frames. Trajectories of features on multiple objects can also
be computed. The algorithm accepts a slow, continuous change of brigh
tness D.C. level in the pixels of the feature. Another important aspec
t of the algorithm is the use of an adaptive feature matching threshol
d that accounts for change in relative brightness of neighboring pixel
s. As applications of the feature tracking algorithm, and to test the
accuracy of the tracking, we show how the algorithm has been used to e
xtract the Focus of Expansion (FOE) and to compute the time-to-contact
using real image sequences of unstructured, unknown environments. In
both applications information from multiple frames is used.