In this thesis, an algorithm for object tracking through frames of video using a fast partial shape matching technique is proposed. The tracking is divided into two modules: 1) moving object extraction followed by color/edge segmentation, and 2) tracking through frames using partial shape matching. The major challenges of object tracking, such as occlusions, splitting of one object and appearance and disappearance of objects, are effectively resolved. The proposed algorithm is tested on several synthetic and real life video sequences and is shown to be very effective in identifying and tracking moving objects independent of translations, rotations, scale variations and occlusions.The novelty of the proposed algorithm lies in its ability to independently track full and partial objects undergoing split, merge and occlusion scenarios independent of their location and scale in the scene. . The technique assumes that: 1) the video frames are captured at 30 frames per second in order for the object(s) motion (translation, rotation, isometric scale variations) to be well modeled by an affine transformation, 2) the object(s) being tracked are larger than a certain number of pixels to allow for comprehensive shape modeling, and 3) the video camera is kept stationary.
展开▼
