Through Wall Human Being Detection Based on Stacked Denoising Auto-encoder Algorithm

摘要

The application of ultra-wideband radar in the detection of through wall human being has been relatively mature. In this paper, the algorithm of Stacked Denoising Auto-encoder (SDAE) is applied to identify and classify the through wall human being status. The unsupervised learning method is used to train the autoencoder network in order to obtain more abstract feature of the original data, and then add a classifier at the end of the network. Use the supervised learning method to fine-tuning the network to get the optimization of the model. Finally, on the network model for testing. Experimental results showed that the Stacked Denoising Auto-encoder deep network can effectively classify and identify the through wall human being status.