Point Target Detection for MAWS Application

Distant aerial threats observed from Missile Approach Warning System (MAWS) appear as point targets therefore point target detection is a significant field of research for the defence sector. In the current work, the authors have carried out an analysis of three distinct combinations of algorithms for point target detection: Top Hat Morphology(THM) with a fixed global threshold, THM with adaptive threshold and THM with corner detection. A number of IR video sequences of sky background are recorded with an ultra wide FOV IR sensor and Gaussian blurred point targets with predefined trajectory have been overlaid on this. In this way a synthetic video data-set has been generated. Authors have used three variants of dataset: 14-bit uncompressed, 8-bit linearly compressed and 8-bit non linearly compressed for performance analysis. Probability of detection and false alarms per frame have been considered as performance measure. Analysis for execution time has also been carried out.