Flying point target tracking using infrared images

Abstract

To improve the detection level and performance of designed infrared motion analysis system, a combined scheme is proposed to get long-term tracklet for flying small objects: 1) point object is efficiently detected via a fast background extraction, and an improved correlation filtering algorithm is utilized for possibly near object with much texture; 2) tracker is initialized and managed by estimating continuous motion using Kalman filter; 3) prior knowledge of object is further incorporated to remove false object in tracklet association process. Outdoor experiment proves the proposed techniques improve the accuracy for target objects, and it also extends the validness of our strategy for coming on orbit system.