Feature selection of hyperspectral data through local correlation and SFFS for crop classification

Abstract

In this paper, we propose a procedure to reduce dimensionality of hyperspectral data while preserving relevant information for posterior crop cover classification. One of the main problems with hyperspectral image processing is the huge amount of data involved. In addition, pattern recognition methods are sensitive to problems associated to high dimensionality feature spaces (referred to as Hughes phenomenon or curse of dimensionality). We propose a dimensionality reduction strategy that eliminates redundant information by means of local correlation criterion between contiguous spectral bands; and a subsequent selection of the most discriminative features based on a Sequential Float Feature Selection algorithm. This method is tested with a crop cover recognition application of six hyperspectral images from the same area acquired with the 128-bands HyMap spectrometer during the DAISEX99 campaign. In the experiments, we analyze the dependence on the dimension and employed metrics. The results obtained using Gaussian Maximum Likelihood improve the classification accuracy and confirm the validity of the proposed approach. Finally, we analyze the selected bands of the input space in order to gain knowledge on the problem and to give a physical interpretation of the results.

Separating Mangrove Species and Conditions Using Laboratory Hyperspectral Data: A Case Study of a Degraded Mangrove Forest of the Mexican Pacific

“…In addition, the large number of wavebands available (i.e., high dimensionality) imposes the problem of high computational cost, the so-called "Hughes phenomenon" or "the curse of dimensionality" [19][20][21][22]. Furthermore, there are typically strong correlations between closely neighboring spectral wavebands.…”

“…However, the steep environmental gradient and the coarse image spectral and spatial resolutions can limit the effectiveness of mapping mangroves [7]. As a consequence, hyperspectral remote sensing techniques, where very large numbers of bands are available (i.e., typically greater than 20), have been shown to be a potential alternative source of data for monitoring wetland vegetation [7], including mangroves [8][9][10]. Spectral responses could be different for various hyperspectral wavebands.…”

Hyperspectral Imaging for Raw Material Sorting and Processed Product Quality Control

“…They work directly with the spectral data space providing advantages such as the interpretability of results. Sequential Floating Feature Selection (SFFS) algorithm (Ferri et al, 1994;Gomez-Chova, 2003) belongs to this category. Fig.…”

Discrimination of Tropical Mangroves at the Species Level with EO-1 Hyperion Data

“…To alleviate this problem, the dimensionality of hyperspectral data needs to be reduced while preserving the key spectral information [63]. In the remote sensing literature, a popular approach to reducing the spectral dimension is to use feature selection algorithms [60,65,67,68,[70][71][72][73]. The genetic search algorithm (GA) is one of the most frequently used band selection found in the recent literature and was also proved to be effective for selecting spectral subsets for vegetation classification [60,72,74].…”

“…This is called the Hughes phenomenon or the curse of dimensionality [67,68]. Furthermore, the co-linearity (i.e., redundant spectral information) also imposes the risk of over fitting when the classification is performed [67,69].…”