High Accuracy Detection and Tracking of Objects

One of the critical tasks in Computer Vision is Detection and Tracking of objects. But still now, the issues related to this are developing. For the automatic detection of moving objects, some of the monitoring systems cannot able to find the difference, when the difference of brightness between the background and the moving objects is small. The costs are very high in these systems. Most of the previous methods, only concentrated on detecting rough area of targets. The accurate moving target detection cannot be achieved. It makes the result to be shown with noise and heals and the computation time is also increased. Many systems are unable to solve critical solutions such as Partial Occlusions and Cross Targets. In my proposed system the Soft Computing techniques can handle objectives and arbitrary constraints with a high degree of simplicity, we use one of the Evolutionary Approach of Soft Computing Technique in this paper. We demonstrate the object detection and tracking methods separately. Initially, a visual object detection approach is presented for minimizing the computation time, in order to achieve high detection accuracy. In this approach, three key contributions are described. Key terms: Video surveillance, Soft computing, object detection, tracking.