Crowd counting method combined with layer aggregation mechanism

The invention discloses a crowd counting method combined with a layer aggregation mechanism, which belongs to the field of computer vision and mainly comprises the following steps of: firstly, preprocessing pictures in a data set to obtain image blocks for network training; the VGG16 backbone network is used to extract low-level features of the image, and dimension reduction is carried out on the number of feature channels so as to reduce the network complexity; then obtaining rich multi-scale information by using five cascaded LAMs, and further optimizing the features through a channel attention model; and finally, generating a crowd density map by using the refined features obtained previously through a back-end network. And adding the pixels in the density map to obtain the total number of people in the image. According to the method, the crowd counting difficulty such as crowd scale change is solved to a certain extent. Experiments show that in scenes and environments containing various crowd densities, t