Dangerous article detection method and device based on cross fusion attention mechanism

The invention discloses a hazardous article detection method based on a cross fusion attention mechanism, and the method comprises the steps: obtaining a terahertz image which comprises a training image and a test image; constructing a deep learning network model, wherein the deep learning network model comprises a backbone feature extraction network, a check feature extraction network, an efficient fusion module, a cross fusion self-attention and a YoloHead detection head; and training the deep learning network model based on the training image to obtain a trained deep learning network model, inputting the test image into the trained deep learning network model, and outputting a hazardous article detection result. According to the invention, an efficient fusion module and cross fusion self-attention are introduced, the feature information in the terahertz image can be effectively utilized, the detection accuracy and robustness of the dangerous goods target are improved, the original detection network is ligh