Comparison of infrared and visible imagery for object tracking: Toward trackers with superior IR performance

Abstract

The subject of this paper is the visual object tracking in infrared (IR) videos. Our contribution is twofold. First, the performance behaviour of the state-of-the-art trackers is investigated via a comparative study using IR-visible band video conjugates, i.e., video pairs captured observing the same scene simultaneously, to identify the IR specific challenges. Second, we propose a novel ensemble based tracking method that is tuned to IR data. The proposed algorithm sequentially constructs and maintains a dynamical ensemble of simple correlators and produces tracking decisions by switching among the ensemble correlators depending on the target appearance in a computationally highly efficient manner. We empirically show that our algorithm significantly outperforms the state-of-the-art trackers in our extensive set of experiments with IR imagery.

Co-difference based object tracking algorithm for infrared videos

“…We compared the proposed co-difference based tracking algorithm with various state-of-the-art trackers: COV [5], TBOOST [2], MILTrack [8], ODFS [9], FCT [10], STRUCK [11], L1APG [12], MOSSE [13], CRC [14] and IVT [15]. All of the above mentioned video object tracking methods are tested on the IR band image sequences of SENSIAC dataset 1 .…”

“…As the surveillance systems started to utilize IR cameras more and more commonly, a need for targeting IR specific challanges has emerged. Even if some recent studies specifically address the issue [2], visual object tracking in IR spectrum,especially with a restricted computational power, presents a challenging task that needs to be studied.…”

“…A more efficient alternative to covariance matrix, the so-called co-difference matrix, was proposed [6] and used in various applications [7]. In this paper, we employ the co-difference matrix in the visual object tracking problem and compare its performance with covariance matrix method as well as other recent state-of-the-art trackers [2,[8][9][10][11][12][13][14][15].…”

Robust Visual Tracking Using Multi-Frame Multi-Feature Joint Modeling

“…On the other hand, [40] employed an online random fern classifier as a re-detection component for long-term tracking, whereas [16] presented a biology-inspired framework where short-term processing and long-term processing are cooperated with each other under a correlation filter framework. Finally, it is worth noting that, with the rapid development of correlation filters on visual tracking, several correlation filter-based thermal infrared trackers (e.g., [41]- [43]) have been developed in recent years. This work only focuses on visual tracking on color video sequences.…”

A thermal Object Tracking benchmark

“…This makes it hard to get an overview of the current status and advances within the field. Furthermore, tracking in thermal infrared video poses different challenges compared to tracking in visual video [22], hence, a separate benchmark is needed. For these reasons, we have prepared and will make publicly available a new thermal infrared benchmark for short-term singleobject (STSO) tracking methods.…”