Geometric Image Analysis Using Multiscale Statistical Features

James M. Coggins and Elizabeth Graves

Many medical image analysis problems involve pixel classification tasks such as segmentation and landmark identification. This paper describes a methodology for defining an optimal feature space for performing such tasks based on a principled, generalized image decomposition. The methodology is demonstrated on two kinds of test images, and the nature of the optimization is illustrated.


This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.