Real-time image segmentation based on learning models

Abstract

This paper presents real-time, digital image segmentation techniques using variable threshold functions. The approach is based on new learning models used to generate the variable threshold functions. The learning models are derived from discrete time functions often used in digital control system design. The techniques are successful to detect regions with different or poor light conditions and can be applied to images with occluded or noisy objects. In addition, the approach can be used to locate objects in a scene. The developed algorithms can also be integrated on a single monolithic integrated circuit or implemented as an embedded system for real-time applications.