WSN and IoT

15Deep Learning Method for Hyperspectral Image Classification

Parul Bhanarkar and Salim Y. Amdani

DOI: 10.1201/9781003437079-15

15.1 Introduction to Hyperspectral Imaging

The collection and analysis of spectrum data from scenes or objects at a finer scale than is achievable with conventional imaging techniques is made possible by advanced remote sensing techniques like hyperspectral imaging (HSI) [1]. Contrary to traditional imaging methods, which only record a small number of distinct spectral bands, hyperspectral imaging captures a far larger number of bands, as illustrated in Figure 15.1. The upshot is that each and every pixel in the image will have a distinct spectral fingerprint [2]. The electromagnetic spectrum is only partially captured ...