Implementation of algorithms for Point target detection and tracking in Infrared image sequences

Abstract

This paper primarily focuses on the study and implementation of two target detection algorithms and also on the formulation of a tracking model to track the detected targets. These detection algorithms- one under the category of Detect before track (DBT) approach and other being Track before detect (TBD) are implemented for point target detection in Infrared Search and Track (IRST) systems for airborne targets. Performance of these algorithms is tested on real Infrared image sequences by plotting receiver operating characteristics (ROC) curves under different scenarios. Results are compared by discussing advantages and shortcomings of both approaches in terms of computational complexity and target detection capability. To track the detected targets with linear trajectories, a tracking algorithm based on Kalman filtering is used. This model is applied for targets under different scenarios and simulation results are presented in this paper.