An Adaptive Approach To Edge Detection Based On Visual Perception
Zhang Tianxu · 2005
In this paper we present a new nonlinear operator for edge detection based on the p rinciples of visual perception. Due to the use of information from image segnentation process, this operator p ossesses both local and global adaptability. Experiments o ver a set of low contrast photographs show that, compared to some linear and shift-invariant operators, t he new detector can result in an entropy increase of t he edge strength, and matches the human performance quite well.