Hyperspectral anomaly detection method based on deep learning network

The invention relates to the technical field of data processing, in particular to a hyperspectral anomaly detection method based on a deep learning network, and the method comprises the steps: S1, obtaining a to-be-processed hyperspectral image; s2, inputting the to-be-processed hyperspectral image into a trained deep learning network to obtain a final anomaly detection result; the trained deep learning network comprises a hidden layer feature extraction network, a deep comprehensive potential feature extraction network, a collaborative representation neural network and a detection fusion network; wherein training data of a hyperspectral image is adopted in advance to carry out joint training on a hidden layer feature extraction network, a deep comprehensive potential feature extraction network, a collaborative representation neural network and a detection fusion network in a deep learning network so as to obtain a trained deep learning network. Compared with the prior art, the detection performance and the d