When a painting is viewed through the predesigned mobile application spectators feel as if the painting gets some life. This sort of a system can be used at art exhibitions. In ARTAR instead of a separate marker being used for image detection, parts of paintings themselves act as markers. When a painting is viewed through the mobile device, These image portions will be detected and are replaced by the frames of predesigned animations. Different keypoint detection and description techniques like SIFT, SURF, ORB, FAST, FREAK, BRISK etc are used to identify the area of interest. The keypoints in the reference image will be initially calculated and these points will be matched with the key points of each frame from camera. If a matching portion is obtained, that particular area is replaced with the desired animation. Also soundscapes can be associated with these areas. The performance of different combinations of detectors and descriptors is analysed and the best combination is used to frame the mobile application.