Abstract
Hyperspectral target detection is widely used in both military and civilian fields. In practical applications, how to select a low-correlation and representative band subset to reduce redundancy is worth discussing. However, most of the existing band selection (BS) methods usually select bands according to the statistics or correlation, which neglect the spectral characteristics of the desired target and are not specially designed for target detection. Therefore, this article proposed a novel BS method, called constrained-target BS with subspace partition (CTSPBS), to select an optimal subset with low internal correlation and strong target representability for the target detection task. By using a specially designed subspace partition method based on correlation distance (CDSP), CTSPBS divides the hyperspectral bands into several unrelated subspaces. Then, according to certain constrained-target band prioritization (BP) criteria, the band with the highest priority in each subset is selected to form the optimal subset for a specific target. Correspondingly, two versions of the proposed method, minimum variance BS with CDSP (CDSP_MinV) and minimum variance BS with CDSP (CDSP_MaxV), are derived to implement CTSPBS. Extensive experiments on three public hyperspectral datasets demonstrate that the proposed method exhibit more robust and effective performance than several state-of-the-art methods. Finally, this article focuses on the difficulty of marine benthos detection in mariculture application and proves the feasibility of the proposed method.
Highlights
H YPERSPECTRAL images (HSIs) have hundreds of contiguous and narrow spectral channels that can provide higher spectral resolution than RGB images [1]
There are a large number of redundant bands or low-signal-to-noise ratio (SNR) bands in HSIs, and some previous studies [21]–[24] have confirmed that the detection of a specific target can be achieved with a small number of representative bands
The constrained-target band prioritization (BP) is derived from the concept of constrained energy minimization (CEM), which was widely used for hyperspectral target detection
Summary
H YPERSPECTRAL images (HSIs) have hundreds of contiguous and narrow spectral channels that can provide higher spectral resolution than RGB images [1]. According to the variance generated by the desired target signal, the latter provides two BP criteria for multiple-target detection, called minimal variance-based band prioritization (MinV-BP) and maximal variance-based band prioritization (MaxV-BP), respectively These above BP-based methods can effectively calculate the band priority sequence for a specific target, they usually regard each band as independent to evaluate their ability to distinguish targets, never treat the hyperspectral bands as ordered, and ignore the strong correlation between adjacent bands. 1) This article proposed a novel BS approach, CTSPBS, for a specific target detection task, which can select an optimal band subset with low internal correlation and strong target representation ability.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have