Hyperspectral Image Classification Based on Dense Pyramidal Convolution and Multi-Feature Fusion

In recent years, hyperspectral image classification techniques have attracted a lot of attention from many scholars because they can be used to model the development different cities and provide reference for urban planning construction. However, due difficulty in obtaining images, only limited number pixels as training samples. Therefore, how adequately extract utilize spatial spectral information images with samples has become difficult problem. To address this issue, we propose method based on dense pyramidal convolution multi-feature fusion (DPCMF). approach, two branches are designed features, respectively. branch, pyramid convolutions non-local blocks multi-scale local global features samples, which then fused obtain features. layers Finally, fed into fully connected results. The experimental results show that overall accuracy (OA) proposed paper is 96.74%, 98.10%, 98.92% 96.67% four datasets, Significant improvements achieved compared five methods SVM, SSRN, FDSSC, DBMA DBDA classification. better exploit when limited. Provide more realistic intuitive terrain environmental conditions planning, design, construction management.