Moving target efficient identification method based on compression observation

The invention discloses a moving target efficient identification method based on compression observation. The objective of the invention is to simultaneously solve two problems, namely low identification accuracy and low calculation efficiency, encountered when a moving target is identified. Firstly, an original vibration signal is linearly projected to a compression domain through a measurement matrix, and a compression observation value is obtained; and then, according to a topological relation and time sequence correlation of the vibration signals in space, a deep neural network used for extracting nonlinear characteristics of a compressed observation value is designed, and through accurate mapping of the observation value and a target category, a high-timeliness and high-accuracy moving target identification method is realized. The method has the beneficial effects that redundant data in original signals can be effectively removed, noise interference is suppressed, and the method has high robustness and a