Mean-Shift Tracking Algorithm for Salient Object Detection in videos

Salient object detection is an interesting subject in the field of video tracking and its applications. The main purpose of salient object detection is to estimate the position of the object in images in a continuous manner and reliably against dynamic scenes. When object is moving then its detection is a challenging task in much vision area. Two tasks are performed to detect salient object, in videos. First detection and second following the object path. This can be beneficially achieved by using the mean shift object tracking algorithm. In this approach, a rectangular target window is defined in an initial frame for a target in a video, and after that, the salient object is separated from the back ground by processing the data within that window. The proposed method has good accuracy to track a moving object in successive frames under some difficulties such as appearance changes due to image noise, etc. We will employ this approach for increasing the capability of moving object tracking.