Adaptive motion estimation and sequential outline separation based moving object detection in video surveillance system

Abstract Moving object detection is an important task in the basic research field of automated video surveillance systems in computer vision and video processing. The application of moving object tracking is critical for military, surveillance systems and operational robot applications, and getting more critical day by day. Closer view gap, ghosting, and sudden lighting changes have been the primary issues in moving object detection in existing methods. To think about the above issues, this work proposes two methodologies like consolidating background subtraction and improved sequential outline separation strategies for the recognition of various moving objects from indoor and outdoor genuine video dataset. This kind of framework can be realized in the general public places such as shopping malls, air terminals and railway stations or the safety of any private premises is the primary concern. This work precisely distinguishes the moving objects with shifting object size and number in various complex situations. The simulation work is done with MATLAB software, to measure the detection error and processing time of the proposed strategy. The proposed sequential outline separation method starts with background subtraction and foreground detection for motion and object discovery and it is a procedure of separating the territory of enthusiasm from developed background. Simulation results and error rate investigation demonstrate that our proposed strategies identify the moving targets productively. As Compared to other conventional systems, our proposed adaptive motion estimation and sequential outline separation method performs better by achieving an accuracy of 97.45%, a sensitivity of 94.2%, and a specificity of 97.72%.