SETh system spatio-temporal object tracking using combined color and motion features

Abstract

This paper presents an overview of an object detection and tracking method used in SETh system (Simple Enemy Tracking). Most of visual systems use only motion features to detect and track objects of interest. Object tracking is a complex problem and thus cannot be solved using only motion data [6]. There are numerous cases where it is required to detect only, for example, small red moving objects. As it is presented, color detection combined with motion detection and tracking managed by history module, allows applications to perform high level object tracking in real-time. Key-Words: Image processing, motion detection, color detection, object tracking, history, real-time