 Hyperspectral imagery has become increasingly popular for its ability to provide detailed spectral information about objects in the environment. This makes it ideal for applications such as military surveillance, mineral exploration and environmental monitoring. However, the large amount of data and spectral variations can make it difficult to detect targets in these images. Several different approaches have been developed to address this challenge, including hypothesis testing, spectral angle, signal decomposition, CM, kernel, sparse representation and deep learning. Each approach has its own strengths and weaknesses, but all share the goal of identifying targets in hyperspectral images. This article was authored by Bowen Chen, Li Qinlu, Zhengxiazou and others.