Video-based Tancho Koi Fish Tracking System Using CSK, DFT, and LOT

Abstract

In this paper, a comparison of Circulant Structure of Tracking-by-detection with Kernels (CSK), Distribution Fields for Tracking (DFT) and Locally Orderless Tracking (LOT) methods is proposed for the process of video-based object tracking. The object in question is the tancho koi fish and its variants, in the form of video recordings. All three methods are applied to trace the red roundabout in the head of the koi fish. The purpose of this study is to provide the best tracking process recommendations for this case. From the experimental results, it is found that the average sample accuracy of the DFT tracking method is more recommended than the other two methods, even though it has an average frame rate that is no better than the CSK method.