Local Binary Patterns and Extreme Learning Machine for Hyperspectral Imagery Classification

By Wei Li, Cheng Chen, Hongjun Su, Qian Du. IEEE Transactions on Geoscience and Remote Sensing, 2015

638 citations recorded by the source corpus.

It is of great interest in exploiting texture information for classification of hyperspectral imagery (HSI) at high spatial resolution. In this paper, a classification paradigm to exploit rich texture information of HSI is proposed. The proposed framework employs local binary patterns (LBPs) to extract local image features, such as edges, corners, and spots. Two levels of fusion (i.e., feature-level fusion and decision-level fusion) are applied to the extracted LBP features along with global Gabor features and original spectral features, where feature-level fusion involves concatenation of multiple features before the pattern classification process while decision-level fusion performs on probability outputs of each individual classification pipeline and soft-decision fusion rule is adopted to merge results from the classifier ensemble. Moreover, the efficient extreme learning machine with a very simple structure is employed as the classifier. Experimental results on several HSI data sets demonstrate that the proposed framework is superior to some traditional alternatives.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/fff45cbbb315fffb4c72d8459fcd9fe28a87519c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

A Real-Time Human Action Recognition System Using Depth and Inertial Sensor Fusion (2016, methodology): “To merge the outcomes of the two classifiers, the logarithmic opinion pool (LOGP) [19] technique is employed.”
Multisource Remote Sensing Data Classification Based on Convolutional Neural Network (2018, methodology): “A network-based classiﬁcation method called extreme learning machine (ELM) [31] has been introduced with comparable performance to SVM. Recently, deep learning-based methods have attracted great attention due to the ability of digging the latent representations and features from the raw data.”
Recent Advances on Spectral–Spatial Hyperspectral Image Classification: An Overview and New Guidelines (2018, methodology): “spatial window to adaptively set neighborhood covering and construct corresponding joint sparse/collaborative representations, while [87] used the central pixels as thresholds to generate adaptive neighborhood covering and accordingly get”
Cascaded Recurrent Neural Networks for Hyperspectral Image Classification (2017, background): “For example, one can directly consider the rich spectral signature as features and feed them into advanced classifiers, such as support vector machine (SVM) [2], random forest [3], and extreme learning machine [4].”
Spectral–Spatial Morphological Attention Transformer for Hyperspectral Image Classification (2023, positive): “Li et al. [24] implemented an ELM to classify HSIs by extracting local binary patterns (LBPs) for classification.”
Land-use scene classification using multi-scale completed local binary patterns (2016, background): “Local binary patterns (LBP) [17] descriptor, as a texture descriptor, has been successfully applied to a variety of imageprocessing andmachinevision applications [18,19].”
Diverse Region-Based CNN for Hyperspectral Image Classification (2018, background): “Meanwhile, sparse representation-based classification (SRC) [15], [16], extreme learning machine (ELM) [17], active learning [18], relevance vector machine (RVM) [19] and other classifiers have been developed to produce superior performance.”
Hyperspectral Image Classification Using Deep Pixel-Pair Features (2017, methodology): “From the results, CNN-PPF still consistently performs better than the three other methods, i.e., SVM-RFS, SVM, ELM, and CNN.”
Deep Learning Meets Hyperspectral Image Analysis: A Multidisciplinary Review (2019, methodology): “This is the case, for instance, of Scale-Invariant Feature Transform (SIFT) [12], Histogram of Oriented Gradients (HOG) [13] or Local Binary Patterns [14], in conjunction with kernel-based Support Vector Machines (SVM) [15], Random Forests [16] or statistical learning methods [17].”
A survey: Deep learning for hyperspectral image classification with few labeled samples (2021, methodology): “Then, a large number of feature extraction methods that integrate the spatial structural and texture information with the spectral features have been developed, including morphological [21, 22, 23], filtering [24, 25], coding [26], etc.”
Deep learning classifiers for hyperspectral imaging: A review (2019, methodology): “…of DNN for HSI classification purposes, for instance Le et al. (2015) review the hyper-parameters used by the spectral and spectral-spatial DBNs of Chen et al. (2015), while Zhong et al. (2017a) present a diversified DBN for HSI data classification, which regularizes the pre-training and…”
Deep Few-Shot Learning for Hyperspectral Image Classification (2019, methodology): “…a large number of spectral–spatial feature extraction methods have been developed for HSI classiﬁcation, such as extended morphological proﬁles [11], Markov random ﬁeld [12], local binary pattern [13], Gabor wavelet ﬁlters [14], discriminant analysis [15], [16], and segmentation approaches [17].”
Multimodal Fusion Transformer for Remote Sensing Image Classification (2022, positive): “, where local binary patterns (LBPs) were extracted from HSIs and used for classification purposes [28].”
PCA-Based Edge-Preserving Features for Hyperspectral Image Classification (2017, methodology): “filtering-based postprocessing method (EPFP) [37], the multiple feature learning (MFL)-based method [49], the local binary patterns (LBPs)-based method [50], the SVM on EMAPs [26], the weighted Markov random fields (WMRFs)-”
Multiscale Dynamic Graph Convolutional Network for Hyperspectral Image Classification (2019, methodology): “In addition to these, graph-based methods [5], extreme learning machine [6], sparse representation-based classiﬁer [7], and many other methods have been further employed to promote the performance of hyperspectral image classiﬁcation.”
Remote Sensing Image Scene Classification Using Multi-Scale Completed Local Binary Patterns and Fisher Vectors (2016, background): “Local binary patterns (LBP) [21,22] are an effective measure of spatial structure information of local image texture.”