Overview

This course provides the knowledge to build image-based decision-making/support systems. Such systems require techniques for representation, description, classification, and segmentation of images.

Representation, description, classification, and segmentation of images are topics with many different techniques. Each topic could easily justify a separate course. Therefore, the course focuses on the main concepts and in the study of successful approaches. Prior knowledge in image processing is important, but the course reviews fundamental concepts and their extension to multidimensional and multiband images.

Image analysis is based on some content of interest, whose representation facilitates measure (feature) extraction. From image descriptors, one can use regression to predict physical attributes, classification to detect events, and segmentation to isolate and quantify objects. The course focuses on image classification and segmentation problems.