Hyperspectral image classification by integrating attention-based LSTM and hybrid spectral networks

AL-Kubaisi, Mohammed Ahmed; Shafri, Helmi Zulhaidi Mohd; Ismail, Mohd Hasmadi; Yusof, Mohd Johari Mohd; Hashim, Shaiful Jahari

doi:10.1080/01431161.2022.2093621

Cited by 6 publications

(8 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…During classification, some popular techniques to optimize the feature distances of hard classes are summarized. [17,20,24,28,31,34,36,38,39,58,61,63,65,67,68,70,72,74,76,77,81,[89][90][91][92]96,106,109,115,119,124,126,128,131,[145][146][147][148]150,152,155,159,164,165,169,176,187,189,190,197,200,…”

Section: Hsi Classification Based On Discriminating Spectral-spatial ...mentioning

confidence: 99%

“…This structural consistence can discover the short dependency between neighboring bands for improving spectral feature representation. As an extension of RNNs, LSTM networks introduce memory units to replace regular neurons, which enhances the ability to sense the spectral dependency between farther bands [65][66][67]. They can acquire both long and short spectral dependencies.…”

Section: Spectral Dependencymentioning

confidence: 99%

“…The SSP models based on decision fusion fuse the respective classification results of spectral and spatial features to predict the final label [67,79,80,157]. As high-level fusion, decision fusion can employ multiple classifiers for different spectral and spatial subnetworks and derives a global decision according to the majority voting rule.…”

Section: Extraction Of Discriminating Spectral-spatial Features 231 A...mentioning

confidence: 99%

“…In contrast, the second one places the spatial subnetwork before the spectral subnetwork [63,65,67,154,210]. The spatial contextual information of samples were compressed into the spectral domain and 1D convolutions or RNNs were used to extract spectral features for classification.…”

Section: Sequential Spatial-spectral (Seq-ae) Modelsmentioning

confidence: 99%

“…In order to classify HSIs, early deep-learning-based models, including artificial neural networks [58][59][60], deep belief networks (DBNs) [61,62], recurrent neural networks (RNNs) [63,64], long short-term memory (LSTM) networks [65][66][67], and auto-encoders (AEs) [68][69][70][71], mainly focused on the spectral features of each spectrum and the flattened spatial structures. Although the spectral dependencies were considered by RNN and LSTM models, the correlations between spectral and spatial information were often disregarded, making the classification accuracy unsatisfying.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Discriminating Spectral–Spatial Feature Extraction for Hyperspectral Image Classification: A Review

Li,

Wang,

Cheikh

2024

Sensors

View full text Add to dashboard Cite

Hyperspectral images (HSIs) contain subtle spectral details and rich spatial contextures of land cover that benefit from developments in spectral imaging and space technology. The classification of HSIs, which aims to allocate an optimal label for each pixel, has broad prospects in the field of remote sensing. However, due to the redundancy between bands and complex spatial structures, the effectiveness of the shallow spectral–spatial features extracted by traditional machine-learning-based methods tends to be unsatisfying. Over recent decades, various methods based on deep learning in the field of computer vision have been proposed to allow for the discrimination of spectral–spatial representations for classification. In this article, the crucial factors to discriminate spectral–spatial features are systematically summarized from the perspectives of feature extraction and feature optimization. For feature extraction, techniques to ensure the discrimination of spectral features, spatial features, and spectral–spatial features are illustrated based on the characteristics of hyperspectral data and the architecture of models. For feature optimization, techniques to adjust the feature distances between classes in the classification space are introduced in detail. Finally, the characteristics and limitations of these techniques and future challenges in facilitating the discrimination of features for HSI classification are also discussed further.

show abstract

Section: Hsi Classification Based On Discriminating Spectral-spatial ...mentioning

confidence: 99%

Section: Spectral Dependencymentioning

confidence: 99%

Section: Extraction Of Discriminating Spectral-spatial Features 231 A...mentioning

confidence: 99%

Section: Sequential Spatial-spectral (Seq-ae) Modelsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Discriminating Spectral–Spatial Feature Extraction for Hyperspectral Image Classification: A Review

Li,

Wang,

Cheikh

2024

Sensors

View full text Add to dashboard Cite

show abstract

A novel model for hyper spectral image enhancement and classification: PCA, MBAO and CNN integration

Lalitha,

Latha

2024

Opt Quant Electron

View full text Add to dashboard Cite

Evaluation method of composite insulator aging status based on hyperspectral imaging technology

Fan,

Guo,

Liu

et al. 2024

Measurement

View full text Add to dashboard Cite

Hyperspectral image classification by integrating attention-based LSTM and hybrid spectral networks

Cited by 6 publications

References 57 publications

Discriminating Spectral–Spatial Feature Extraction for Hyperspectral Image Classification: A Review

Discriminating Spectral–Spatial Feature Extraction for Hyperspectral Image Classification: A Review

A novel model for hyper spectral image enhancement and classification: PCA, MBAO and CNN integration

Evaluation method of composite insulator aging status based on hyperspectral imaging technology

Contact Info

Product

Resources

About