Laser & Optoelectronics Progress, Volume. 60, Issue 12, 1210005(2023)
Hyperspectral Image Classification Combining Superpixel Principal Component Analysis Dimensionality Reduction with Extended Random Walk Probability Optimization
Fig. 1. Hyperspectral image classification framework based on SE_SVM
Fig. 2. False-color image (bands 34, 17, 10) and ground truth label map of Indian Pines dataset
Fig. 3. False-color image (bands 68, 27, 19) and ground truth label map of Pavia University dataset
Fig. 4. False-color image (bands 68, 27, 19) and ground truth label map of Salinas dataset
Fig. 5. Line graphs of overall classification accuracy of hyperspectral dataset with the number of principal components and the number of superpixels
Fig. 6. Classification result graphs of seven methods on Indian Pines dataset. (a) Ground truth label map; (b) SVM; (c) PCA_SVM; (d) SPCA_SVM; (e) ERW_SVM; (f) 3DCNN; (g) SSRN; (h) SE_SVM
Fig. 7. Classification result graphs of seven methods on the Pavia University dataset. (a) Ground truth label map; (b) SVM; (c) PCA_SVM;(d) SPCA_SVM; (e) ERW_SVM; (f) 3DCNN; (g) SSRN; (h) SE_SVM
Fig. 8. Classification result graphs of seven methods on the Salinas dataset. (a) Ground truth label map; (b) SVM; (c) PCA_SVM; (d) SPCA_SVM; (e) ERW_SVM; (f) 3DCNN; (g) SSRN; (h) SE_SVM
|
|
|
|
|
|
|
Get Citation
Copy Citation Text
Dejia Hu, Yuan Huang, Bin Yang, Xinguang He. Hyperspectral Image Classification Combining Superpixel Principal Component Analysis Dimensionality Reduction with Extended Random Walk Probability Optimization[J]. Laser & Optoelectronics Progress, 2023, 60(12): 1210005
Category: Image Processing
Received: Jan. 25, 2022
Accepted: Jun. 14, 2022
Published Online: Jun. 5, 2023
The Author Email: Yang Bin (yangbin@hunnu.edu.cn)