On Hyperspectral Image Classification and Deep Learning
摘要
Hyperspectral imaging is a technique for capturing images of objects in multiple dimensions. It combines spectroscopic and imaging technology to produce multidimensional images. Hyperspectral imaging (HSI) allows us to examine any object's internal and external characteristics. Every characteristic of an object has a distinctive spectral signature that is created as a result of variations in the reflectance or emittance of the object's material. Hyperspectral imaging (HSI), which is non-destructive, is now being used in the food production, medical diagnosis, precision agriculture, pharmaceutical, recycling, and environmental monitoring industries. The various HSI classification techniques based on conventional, deep learning, and pre-trained classifiers will be reviewed.