Abstract: Detecting and tracking an infrared (IR) aerial target in the event of infrared jamming is still challenging. In this paper, singular value decomposition (SVD) based algorithms are proposed for the detection and tracking of aerial targets in the presence of decoy flares. Features including brightness intensity, dominant orientation and complexity were used for the discrimination against decoy flares. SVD characteristics including oriented energy analysis, image energy truncation, and low rank approximation were used to measure these features.