Object Recognition System Based on Oriented Fast and Rotated Brief

Abstract

Over the last few years, object recognition systems take the researchers' attention. The revolution in the artificial intelligence and computer vision systems and algorithms drives to numerous approaches. In this research, we are presenting an object recognition system that used ORB (Oriented FAST and Rotated BRIEF) as a feature extraction method, the proposed approach used as a robot system that performs objects detection and recognition tasks. We used ORB to detect preset features in a given video frame obtained from the camera at real-time and then used Brute-force matcher to match a new video frame features with the object image that we want the robot to track it and recognize it. The matched point must have a distance ratio of 0.65.