Implementation of efficient human object video tracking methods in cluttered environment

Abstract

Tracking of human motion and their body parts under dynamic environment has been the difficult task during video detection situations. On the contrary, considering the movement of the targeted object and occlusion problem is unavoidable while tracking. Also, the tracking problem increases in the crowded and cluttered environmental conditions because of the objects under rotation and changes in illumination. The aim of the research work is to propose the human object video tracking method to estimate the movement of human upper body actions and classify them under dynamic environment. Hence an efficient method to overcome the adverse effects like occlusion upon object tracking is needed. So, the need for this proposal is to develop the contour based dynamic human object tracking algorithm with great accuracy under cluttered environment and occlusions. The dynamic human object tracking algorithm analyses various human object tracking parameters. This was implemented based on the tracking of human objects obtained from the video sequence in static and moving conditions, to explain the greater accuracy under dynamic environment. The algorithm uses circular thresholding for segmentation instead of traditional linear thresholding which finds the threshold value on rotation basis according to the input variations to deliver more accurate detection results. Further improvement in tracking scenario, an efficient and effective newlinetracking of crowded video object sequence with frequently changing background and fast movements of objects are considered. Micro level analysis is considered during the crowded behaviour detection in the video sequence for the movement of the targeted objects. newline newline