Foreground-Background Discriminant Scale Tracker

Muhammad Shehzad Hanif¹, Shafiq Ahmad², Khurram Khurshid²

¹Department of Electrical and Computer Engineering, King Abdulaziz University, Jeddah, Kingdom of Saudi Arabia

²Department of Electrical Engineering, Institute of Space Technology, Islamabad, Pakistan

Abstract

In this article, we propose a tracker two kinds of improvements to a baseline tracker that employs the tracking by detection framework. First, we explore different feature spaces by employing features commonly used in object detection to improve the performance of detector in feature space. Second, we propose a robust scale estimation algorithm that estimates the size of the object in the current frame. Our experimental results on the challenging OTB-13 dataset show that reduced dimensionality Histogram of Oriented Gradients (PCA-HoG) boosts the performance of the tracker. The proposed scale estimation algorithm provides a significant gain and reduces the failure of the tracker in challenging scenarios. The improved tracker called Foreground-Background Discriminant Scale Tracker (FBDST) is compared with 13 state of the art trackers. The quantitative and qualitative results show that the performance of the tracker is comparable to the state of the art against initialization errors, variations in illumination, scale and motion, out-of-plane and in-plane rotations, deformations and low resolution.

Downloads

Full Paper: On the improvement of foreground–background model-based object tracker

Supplementary material for IET-CV submission (see below)

Our results on OTB-13 dataset (see below)