A Review of Real Time Image Processing for Object Detection

In past days, capture images with very high quality and good size is so easy because of rapid improvement in quality of capturing device with less costly but superior technology. Videos are a collect of sequential images with a constant time interval. So video can provide also more information about our object when scenarios about to changing with respect to time. Therefore, manually handling videosit can be quite impossible. That time all that need an automatic devise to process these videos. In this thesis one such attempt has been made to track objects in videos. Many algorithms and technology have been developed to automate monitoring the object in a video file. Object detection and tracking is a one of the challenging task in computer vision. Mainly there are three basic steps in video analysis: Detection of objects of Interest from moving objects, Tracking of that interested objects in consecutive frames, and Analysis of object tracks to understand their behavior Some common choice to choose suitable feature to categories, visual objects are intensity, shape, color and feature points. In this thesis, we studied about mean shift tracking based on the color pdf, optical flow tracking based on the intensity and motion; SIFT tracking based on scale invariant local feature points.