Feature Descriptors for Tracking by Detection: a Benchmark
This work addresses the lack of evaluation for descriptors in tracking applications, providing a benchmark for computer vision researchers, but it is incremental as it applies existing methods to a new context.
The paper benchmarks local feature descriptors for tracking by detection, finding that binary descriptors like ORB and BRISK achieve comparable performance to SIFT or AKAZE in terms of distinctiveness, precision, and speed, due to a higher number of key-points.
In this paper, we provide an extensive evaluation of the performance of local descriptors for tracking applications. Many different descriptors have been proposed in the literature for a wide range of application in computer vision such as object recognition and 3D reconstruction. More recently, due to fast key-point detectors, local image features can be used in online tracking frameworks. However, while much effort has been spent on evaluating their performance in terms of distinctiveness and robustness to image transformations, very little has been done in the contest of tracking. Our evaluation is performed in terms of distinctiveness, tracking precision and tracking speed. Our results show that binary descriptors like ORB or BRISK have comparable results to SIFT or AKAZE due to a higher number of key-points.