EXNet: (2+1)D Extreme Xception Net for Hyperspectral Image Classification

Abstract

3D-CNNs have demonstrated their capability to capture intricate non-linear relationships within Hyperspectral Images (HSIs). However, the computational complexity of 3D CNNs often leads to slower processing speeds, limited generalization, and susceptibility to overfitting. In response to these challenges, this study introduces the concept of depthwise separable convolutions using (2+1)D convolutions as an alternative to traditional 3D convolutions for Hyperspectral Image Classification (HSIC). The study observes that (2+1)D convolutions can effectively approximate the complex relationships represented by 3D convolutions while requiring fewer convolutional operations, thereby reducing the computational overhead associated with classification. Experimental results obtained from benchmark HSI datasets, including Indian Pines, Botswana, Pavia University, and Salinas, demonstrate that the proposed model yields results that are comparable to those achieved by various state-of-the-art models in the existing literature. The source code is available on GitHub github.com/mahmad00/Extreme-Xception-Net-for-HSIC.

Damage Detection in Glass Fibre Composites Using Cointegrated Hyperspectral Images

“…The current methods developed for overcoming the lighting issue in vision-based methods, and especially in hyperspectral imaging, have been advanced classification methods, such as classifiers operating on spectral–spatial data [ 18 , 19 , 20 ], where instead of classifying each pixel alone, the data about the neighbouring pixels and location of the pixel in question are also included in the learning data [ 21 ], which lowers the method’s sensitivity to imperfect lighting. This method, however, performs best in scenarios where different endmembers (classes) present on the hyperspectral image are of a different chemical composition from one another, resulting in a high separation of those classes.…”