After the extraction of feature points and descriptors from reference image and camera frame, a similarity check must be carried out to find whether the reference image is present in the camera frame. Since the BRISK descriptors obtained are binary in nature, hamming distance can be used to find similarity. The method of the K-nearest neighbor (K-NN) matching is used to find the best k matching feature points in camera frame corresponding to the reference image. After the matches are found, Lowe’s ratio test is carried out to eliminate error matching. Fig.4 shows the key point matching between the reference image and camera frame. Reference image is at the left side and video frame at right. The matching key points are connected using lines.