The proposed ARTAR algorithm for adding animations and sounds into paintings and other artworks is shown in fig.2. Input video is converted into frames and is analysed to see whether the reference image is present in it or not. Animation is designed corresponding to a reference portion of the image and then the feature points and their descriptors are found in the reference image and the frames from camera. Then these feature descriptors will be matched using similarity measures to find the matching feature points. When feature matching is obtained, homography matrix is calculated to accurately place the desired video frame over the camera frame.