Abstract

Hyperspectral imaging has emerged as a powerful tool for identifying and remotely sensing complex materials in a wide range of applications, including medical diagnostics, security, food safety, and precision agriculture [1,2]. Despite significant advances in this field, there remain a number of challenges that must be addressed in order to fully realize the potential of hyperspectral imaging in real-world applications. One major challenge is the high cost and slow acquisition time of current state-of-the-art hyperspectral imaging systems, which can exceed 20.000 USD for a single camera and take up to a minute to acquire a single image[3]. Additionally, existing systems are often limited by low spatial resolution and require large amounts of memory storage.

© 2023 IEEE