Object Tracking in Hyperspectral Videos with Convolutional Features and Kernelized Correlation Filter

TL;DR

This paper introduces a novel hyperspectral object tracking method combining convolutional features with Kernelized Correlation Filter, effectively encoding spectral-spatial information and outperforming existing methods on hyperspectral videos.

Contribution

It proposes a simple two-layer convolutional network for spectral-spatial feature extraction in hyperspectral videos, eliminating the need for extensive offline training.

Findings

Outperforms several state-of-the-art methods on hyperspectral videos

Effective encoding of spectral-spatial information with a simple network

Robust tracking demonstrated in experimental results

Abstract

Target tracking in hyperspectral videos is a new research topic. In this paper, a novel method based on convolutional network and Kernelized Correlation Filter (KCF) framework is presented for tracking objects of interest in hyperspectral videos. We extract a set of normalized three-dimensional cubes from the target region as fixed convolution filters which contain spectral information surrounding a target. The feature maps generated by convolutional operations are combined to form a three-dimensional representation of an object, thereby providing effective encoding of local spectral-spatial information. We show that a simple two-layer convolutional networks is sufficient to learn robust representations without the need of offline training with a large dataset. In the tracking step, KCF is adopted to distinguish targets from neighboring environment. Experimental results demonstrate that the proposed method performs well on sample hyperspectral videos, and outperforms several state-of-the-art methods tested on grayscale and color videos in the same scene.