Detection of spot target in infrared clutter with morphological filter

The morphological filters are a class of nonlinear signal processing algorithms, which have been applied extensively to computer vision, image processing, and more recently target detection. This paper presents a morphology-based algorithm for spot target detection in IR clutter. The algorithm utilizes a prior knowledge of target size to suppress background clutter and remove noise. The experimental results show that the approach can detect spot target in the presence of clutter for low signal-to-noise ratio (SNR). The probability of detection and false alarm is presented as ROC curves for sky, ground and sea clutter background.