A novel feature recognition algorithm based on binboost

With the high real-time performance and small memory occupancy prevails, the ORB feature has been widely used for image feature matching on mobile devices. However, considering that the existing ORB´s inability of features discrimination and easily causes large of error matching, we propose a novel feature recognition algorithm based on binboost. Firstly, AGAST is used to speed up the detecting process of feature points. Secondly, the attribute of both rotational and scale invariance is constructed on it as well. Finally, the experimental results show that the novel algorithm can achieve encouraging performance of descriptor distinguish while keeping the real-time efficiency of ORB, which proved the effectiveness of our method and its adaptability to mobile devices.