Analysis of feature detector and descriptor combinations with a localization experiment for various performance metrics

Authors : ERTUĞRUL BAYRAKTAR, PINAR BOYRAZ

Pages : 2444-2454

View : 18 | Download : 10

Publication Date : 0000-00-00

Article Type : Research Paper

Abstract :The purpose of this study is to provide a detailed performance comparison of feature detector/descriptor methods, particularly when their various combinations are used for image-matching. The localization experiments of a mobile robot in an indoor environment are presented as a case study. In these experiments, 3090 query images and 127 dataset images were used. This study includes five methods for feature detectors insert ignore into journalissuearticles values(features from accelerated segment test insert ignore into journalissuearticles values(FAST);, oriented FAST and rotated binary robust independent elementary features insert ignore into journalissuearticles values(BRIEF); insert ignore into journalissuearticles values(ORB);, speeded-up robust features insert ignore into journalissuearticles values(SURF);, scale invariant feature transform insert ignore into journalissuearticles values(SIFT);, and binary robust invariant scalable keypoints insert ignore into journalissuearticles values(BRISK);); and five other methods for feature descriptors insert ignore into journalissuearticles values(BRIEF, BRISK, SIFT, SURF, and ORB);. These methods were used in 23 different combinations and it was possible to obtain meaningful and consistent comparison results using the performance criteria defined in this study. All of these methods were used independently and separately from each other as either feature detector or descriptor. The performance analysis shows the discriminative power of various combinations of detector and descriptor methods. The analysis is completed using five parameters: insert ignore into journalissuearticles values(i); accuracy, insert ignore into journalissuearticles values(ii); time, insert ignore into journalissuearticles values(iii); angle difference between keypoints, insert ignore into journalissuearticles values(iv); number of correct matches, and insert ignore into journalissuearticles values(v); distance between correctly matched keypoints. In a range of ${60^{\circ}}$, covering five rotational pose points for our system, the FAST-SURF combination had the lowest distance and angle difference values and the highest number of matched keypoints. SIFT-SURF was the most accurate combination with a 98.41% correct classification rate. The fastest algorithm was ORB-BRIEF, with a total running time of 21,303.30 s to match 560 images captured during motion with 127 dataset images.
Keywords : Feature detectors and descriptors, performance evaluation, performance metrics, localization, object matching