Image Denoising, Deblurring and Object Tracking

Noise reduction is an important step for development of any sophisticated algorithms in computer vision and image processing. A tradeoff between the removed noise and the blur in the image always exists. The capability of the wavelets to give detail spatial-frequency information is the main reason for the use of wavelets. This property promises a possibility for better discrimination between the noise and the real data. Successful exploitation of the wavelet transform might reduce the blurring effect or even overcome it completely. Object tracking is a problem of estimating the positions and other relevant information of moving objects in the sequences of image video. The main difficulties in reliable tracking of moving objects include: rapid appearance changes caused by image noise and interaction between multiple objects. In a long sequence of image video, if the dynamics of the moving object is known, prediction can be made about the positions of the object in a particular frame. This information can be combined with the actual image observation to achieve more robust results. We have explored a possibility to use wavelet transform for object tracking.