Authors : Shahad Mahgoob Nafi and Sawsen Abdulhadi Mahmood
Abstract: Motion detection in video sequences is considered as a critical issue in video surveillance system. the key result of motion detection process is basically aims to extract and track the moving objects by segmenting the original frame into foreground part (moving objects) and background part in a video sequences. Practically, the moving objects may exhibit abnormal actions such as steady persons for while then moving. Meantime, the background may include periodic local actions such as waving arbor. Such intricacies may cause incorrect detection of moving objects based on the existing background subtraction approaches. In this study, we present a modified approach for moving objects detection based on Bhattacharyya distance measurement to detect the interaction between 2 sequence frames in a video sequences captured by static camera taking in consideration an indoor and outdoor scenes. A modified background subtraction method is adopted to extract the foreground parts in the video sequences (moving objects). The performance evaluation of the proposed approach was conducted using different video files taken from public video database with assist of several performance metrics. The experiments results proved the effectiveness of the proposed approach for moving objects detection.
Shahad Mahgoob Nafi and Sawsen Abdulhadi Mahmood, 2019. Moving Objects Detection Based on Bhattacharyya Distance Measurement. Journal of Engineering and Applied Sciences, 14: 4043-4051.