MYSTERIOUS ACTION DETECTION BY SPEED AND DISPLACEMENT ANALYSIS IN IMAGE PROCESSING

There are now three moving target detection algorithms: The optical flow methodology, the background difference technique, and the inter-frame difference method are all employed. The optical flow technique has the most complex algorithm and takes the longest to process data compared to other approaches. The background subtraction(BS) approach, on the other hand, is recognised for its high sensitivity to light changes. Finally, while the frame difference approach is very straightforward and quick to use, the findings are insufficiently accurate due to backdrop brightness fluctuations causing mistakes. In this paper, the objective is to discover strange actions in image processing using speed and displacement analysis. Motion is detected using Incremental and Multi-feature Tensor Subspace Learning (IMTSL) Applied for Background Modeling and Subtraction. Following IMTSL Background Subtraction, different image analysis were performed on the resulting pictures for displacement estimate, followed by speed computation. The results demonstrate improved backdrop modelling and action prediction.