A rule-based soft-decision method for change detection

Because of error propagation, change is often overestimated by the traditional hard-decision method. Aiming at the characteristics of land use change at the urban fringe areas and the research focus into account, this paper proposes a rule-based soft-decision method for change detection. Based on the pixel status of change/unchange, land use change trajectories, and the spatial features of detected changes which are shape, size and location, this method evaluates the rationality of the detected changes. It reaches higher change detection accuracy by employing a series of logic rules in turn in the comparison of post-classification results and by distinguishing between the real land use changes and the potential classification errors. Experimental results prove that the rule-based soft-decision method has better performance than the traditional hard-decision method, which reduces errors caused by misclassification, decreases occurs of pseudo-changes, and improves the change detection accuracy significantly.