Journal of Infrared and Millimeter Waves, Volume. 40, Issue 3, 400(2021)
Hyperspectral image classification combing local binary patterns and k-nearest neighbors algorithm
Fig. 1. (a) LBPs=4,r=1,(b) LBPs=8,r=1,(c) schematic diagram of LBP
Fig. 2. Flowchart of hyperspectral image classification based on the LBP-SSKNN
Fig. 3. (a)False-color composite image(b)ground-truth classes for the Pavia University scene
Fig. 4. (a)False-color composite image(b) ground truth classes for the Indian Pines scene
Fig. 5. False-color composite image(a) and ground truth classes,(b) for the Salinas scene
Fig. 6. Influence on classification accuracies for the number of principal components
Fig. 7. Influence on classification accuracies of the Indian Pines dataset for r and s
Fig. 8. Influence on classification accuracies of the Indian Pines dataset for k
Fig. 9. Classification maps using the four methods for the Pavia University dataset
Fig. 10. Classification maps using the four methods for the Indian Pines dataset
Fig. 11. Classification maps using the four methods for the Salinas dataset
Fig. 12. Classification accuracies under different training samples for the Pavia University dataset
Fig. 13. Classification accuracies under different training samples for the Indian Pines dataset
Fig. 14. Classification accuracies under different training samples for the Salinas dataset
|
|
|
|
|
|
|
|
Get Citation
Copy Citation Text
Jin-Ling ZHAO, Lei HU, Hao YAN, Guo-Min CHU, Yan FANG, Lin-Sheng HUANG. Hyperspectral image classification combing local binary patterns and k-nearest neighbors algorithm[J]. Journal of Infrared and Millimeter Waves, 2021, 40(3): 400
Category: Research Articles
Received: Jun. 29, 2020
Accepted: --
Published Online: Sep. 9, 2021
The Author Email: Jin-Ling ZHAO (zhaojl@ahu.edu.cn), Lin-Sheng HUANG (linsheng0808@163.com)