Hyperspectral Image Classification Based on Automatic Threshold Attribute Profiles and Spatial-Spectral Encoding Union Features

Fig. 6. Filtered images. (a) Third filtering; (b) second filtering; (c) first filtering; (d) first PC; (e) first filtering; (f) second filtering; (g) third filtering

Download full size

Fig. 7. Accuracy comparison results of different methods on 9 categories for Pavia University dataset

Download full size

Fig. 8. Accuracy comparison results of different methods on 16 categories for Salinas dataset

Download full size

Table 1. Attribute and threhold
View table
Table 1. Attribute and threhold
Dataset Pavia University Salinas
Area 55879，93720，131561，169402 24174，45732，67290，88848
Standard deviation 14，26，39，52 12，22，32，42

Table 2. Parameter settings of different datasets
View table
Table 2. Parameter settings of different datasets
Dataset $λ$ Number of hidden layer neurons S-SFMSAE structure
Pavia University 0.1 80 171-100-80-9
Salinas 0.001 80 255-120-80-16

Table 3. Comparison of various methods on Pavia University
View table
Table 3. Comparison of various methods on Pavia University
Dataset S-SFM SAE CDA SAE SVM-RFS 1-D CNN CNN-PPF LBP-ELM
OA 99.28 97.59 97.59 91.10 92.27 96.48
AA 99.01 97.66 92.92 93.30 96.98 91.81
Kappa 99.05 96.86 96.90 88.53 89.89 95.48

Table 4. Comparison of various methods on Salinas
View table
Table 4. Comparison of various methods on Salinas
Dataset S-SFM SAE CDA SAE SVM-RFS 1-D CNN CNN-PPF LBP-ELM
OA 98.32 96.07 93.15 89.28 94.80 92.42
AA 98.91 97.56 96.87 94.83 97.73 96.31
Kappa 98.13 96.78 92.35 88.13 94.17 91.55

Tools

Get Citation

Copy Citation Text

Peiqi Yang, Mingjun Wang. Hyperspectral Image Classification Based on Automatic Threshold Attribute Profiles and Spatial-Spectral Encoding Union Features[J]. Laser & Optoelectronics Progress, 2023, 60(12): 1210016

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites