Abstract
The wavelet (packet) transform has been widely used for texture analysis; however, the extracted features of similar textures with symmetric orientations are indistinguishable. Motivated by the AM-FM representation, the so called modulated wavelet (packet) transform that can be implemented efficiently by the conventional pyramid (tree) structured algorithms is developed. The performance of this new transform is demonstrated on the segmentation of Brodatz textures and an aerial image of San Francisco.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have