Automatic Target Detection in Cluttered Infrared Imagery using Background Modeling Approach
Yash Khare, T Vishwaak Chandran, Abhijit Ramesh, Akhil KG, Lekha S. Nair · 2022 6th International Conference on Trends in Electronics and Informatics (ICOEI) · 2022
Target detection is the primary purpose of a real-time visual surveillance system. However, this is a challenging problem due to high variability in the appearance of targets and the computational cost of existing algorithms prevents real-time deployment. This study proposes a method for real-time automatic detection of moving ground targets such as, vehicles, humans, animals, etc., in image sequences captured by a stationary infrared imaging system. The proposed research work utilizes a background modeling approach to reduce the computation cost when compared to existing algorithms in dynamic background conditions (when the camera itself is moving). The proposed method uses statistical variations of pixels over the temporal domain to compute the background model. Experimental results demonstrate that the proposed algorithm can detect intruding targets from infrared imaging video with a detection sensitivity of 0.88 and a false alarm rate of 0.001.