USING MODELS OF FEATURE PERCEPTION IN DISTORTION MEASURE GUIDANCE

In this paper we present three error measures based on feature percept ion models, in which pixel errors are computed on locations at which h umans might perceive features in the reference image. In the first par t of this work, the three schemes of feature detection will be discuss ed and evaluated in terms of their performance for a simple visual sig nal-processing task. The first model is based on the use of local inte nsity gradients, the second based on the use of phase congruency in an image, and the third based on the use of local energy maxima for a fe w active sensors under a multichannel organization of the reference pi cture. In the second part of this paper, examples are provided of obje ct detection and recognition applications that illustrate the ability of the induced error measures to predict the detectability of objects in natural backgrounds as well as their perceptual capabilities. (C) 1 998 Elsevier Science B.V.