Classification of hyperspectral data using extended attribute profiles based on supervised and unsupervised feature extraction techniques

Marpu, Prashanth Reddy; Pedergnana, Mattia; Mura, Mauro Dalla; Peeters, Stijn; Benediktsson, Jón Atli; Bruzzone, Lorenzo

doi:10.1080/19479832.2012.702687

Cited by 56 publications

(46 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, some novel classification strategies have proposed for classifying hyperspectral images, such as random forest [35]. However, to evaluate the robustness of the deep features, the widely-used SVM classifier is considered as the benchmark in the experiment.…”

Section: Comparison Of Different Methodsmentioning

confidence: 99%

Hyperspectral Imagery Classification Using Sparse Representations of Convolutional Neural Network Features

Liang

2016

Remote Sensing

210

View full text Add to dashboard Cite

Abstract:In recent years, deep learning has been widely studied for remote sensing image analysis. In this paper, we propose a method for remotely-sensed image classification by using sparse representation of deep learning features. Specifically, we use convolutional neural networks (CNN) to extract deep features from high levels of the image data. Deep features provide high level spatial information created by hierarchical structures. Although the deep features may have high dimensionality, they lie in class-dependent sub-spaces or sub-manifolds. We investigate the characteristics of deep features by using a sparse representation classification framework. The experimental results reveal that the proposed method exploits the inherent low-dimensional structure of the deep features to provide better classification results as compared to the results obtained by widely-used feature exploration algorithms, such as the extended morphological attribute profiles (EMAPs) and sparse coding (SC).

show abstract

Section: Comparison Of Different Methodsmentioning

confidence: 99%

Hyperspectral Imagery Classification Using Sparse Representations of Convolutional Neural Network Features

Liang

2016

Remote Sensing

210

View full text Add to dashboard Cite

show abstract

“…The well-known transformation-based characteristics learning methods include principal component analysis (PCA), minimum noise fraction (MNF), etc. PCA [32] can express data in minimum mean square error, but it will be influenced by noise. Therefore, Green et al [30] and Lee et al [31] proposed minimum noise separation methods, which arrange the components of the transformation according to the order of signal-to-noise ratio (SNR).…”

Section: Related Workmentioning

confidence: 99%

“…Transformation-based feature learning methods [29][30][31][32] map or transfer the original data from the high-dimensional data space into the low-dimensional feature space. The well-known transformation-based characteristics learning methods include principal component analysis (PCA), minimum noise fraction (MNF), etc.…”

Section: Related Workmentioning

confidence: 99%

When Low Rank Representation Based Hyperspectral Imagery Classification Meets Segmented Stacked Denoising Auto-Encoder Based Spatial-Spectral Feature

2018

View full text Add to dashboard Cite

When confronted with limited labelled samples, most studies adopt an unsupervised feature learning scheme and incorporate the extracted features into a traditional classifier (e.g., support vector machine, SVM) to deal with hyperspectral imagery classification. However, these methods have limitations in generalizing well in challenging cases due to the limited representative capacity of the shallow feature learning model, as well as the insufficient robustness of the classifier which only depends on the supervision of labelled samples. To address these two problems simultaneously, we present an effective low-rank representation-based classification framework for hyperspectral imagery. In particular, a novel unsupervised segmented stacked denoising auto-encoder-based feature learning model is proposed to depict the spatial-spectral characteristics of each pixel in the imagery with deep hierarchical structure. With the extracted features, a low-rank representation based robust classifier is then developed which takes advantage of both the supervision provided by labelled samples and unsupervised correlation (e.g., intra-class similarity and inter-class dissimilarity, etc.) among those unlabelled samples. Both the deep unsupervised feature learning and the robust classifier benefit, improving the classification accuracy with limited labelled samples. Extensive experiments on hyperspectral imagery classification demonstrate the effectiveness of the proposed framework.

show abstract

“…As a result, we can perform dimensionality reduction, using techniques such as PCA [14], and then perform attribute filtering on the first few PCs, as suggested in [27], in order to reduce computational complexity. For multispectral data, since there are only a few spectral bands available, we can perform attribute filtering on the full original spectral data.…”

Section: A Emapsmentioning

confidence: 99%

“…Although classifiers such as support vector machines (SVMs) [20], with reduced sensitivity to limited training samples, have been successfully used in the literature [6], [21]- [26], the information extracted by morphological approaches can be made more compact in order to further optimize the classification process. This issue was investigated in [16], [27], and [28] by applying several feature extraction and selection techniques to morphological profiles and APs prior to classification, aiming at reducing the high dimensionality of the profile by keeping only (few) relevant features. A compact representation of the profile was also proposed in [13] by defining the so-called morphological characteristic (MC), which derives from the DMP if the underlying region of each pixel is mostly darker or brighter with respect to its surroundings.…”

mentioning

confidence: 99%

Remotely Sensed Image Classification Using Sparse Representations of Morphological Attribute Profiles

Song¹,

Li²,

Mura³

et al. 2014

IEEE Trans. Geosci. Remote Sensing

160

View full text Add to dashboard Cite

Abstract-In recent years, sparse representations have been widely studied in the context of remote sensing image analysis. In this paper, we propose to exploit sparse representations of morphological attribute profiles for remotely sensed image classification. Specifically, we use extended multiattribute profiles (EMAPs) to integrate the spatial and spectral information contained in the data. EMAPs provide a multilevel characterization of an image created by the sequential application of morphological attribute filters that can be used to model different kinds of structural information. Although the EMAPs' feature vectors may have high dimensionality, they lie in class-dependent low-dimensional subpaces or submanifolds. In this paper, we use the sparse representation classification framework to exploit this characteristic of the EMAPs. In short, by gathering representative samples of the low-dimensional class-dependent structures, any given sample may by sparsely represented, and thus classified, with respect to the gathered samples. Our experiments reveal that the proposed approach exploits the inherent low-dimensional structure of the EMAPs to provide state-of-the-art classification results for different multi/hyperspectral data sets.

show abstract

Classification of hyperspectral data using extended attribute profiles based on supervised and unsupervised feature extraction techniques

Cited by 56 publications

References 26 publications

Hyperspectral Imagery Classification Using Sparse Representations of Convolutional Neural Network Features

Hyperspectral Imagery Classification Using Sparse Representations of Convolutional Neural Network Features

When Low Rank Representation Based Hyperspectral Imagery Classification Meets Segmented Stacked Denoising Auto-Encoder Based Spatial-Spectral Feature

Remotely Sensed Image Classification Using Sparse Representations of Morphological Attribute Profiles

Contact Info

Product

Resources

About