Abstract

Scale-space is an important recent concept used in image processing and pattern recognition. Traditional scale-space is generated by a linear smoothing operation. We present here a nonlinear type of smoother related to mathematical morphology which meets (modified) "scale-space axioms " and also generates a "scale-space." This scale-space is full-plane and preserves the positions of features. 1 Introduction At the 7th IJCAI in 1983, Andrew Witkin introduced the concept of scale-space filtering in a, now classical, four page paper [9]. It had been known for some time that the human visual system processes images via a number of spatialfrequency -tuned channels and that a similar multiscale approach is useful in computer vision [7]. However, the question of how to combine signal descriptions at different scales remained a problem. Enter Witkin with the intuitively elegant idea to use a continuous scale parameter for the spatial filter; then the signal features (in this case zero-crossin...

Citations