The embodiment of the invention discloses a pedestrian re-identification method based on compact ternary loss, and the method comprises the steps: inputting a training set pedestrian image into a pre-training deep learning model, and obtaining a first pedestrian image feature; calculating the compact ternary loss of the first pedestrian image feature according to the compact ternary loss function;normalizing the first pedestrian image features to obtain second pedestrian image features; calculating the cross entropy loss of the second pedestrian image feature according to the cross entropy loss function; optimizing a pedestrian re-identification network framework in combination with the compact ternary loss and the cross entropy loss; and performing pedestrian identification based on theoptimized pedestrian re-identification network framework. According to the method, the advantages of compact ternary loss and cross entropy loss are fully utilized, the two kinds of loss are combinedto execute multi-task operation, pedestrian features are learned, and the matching accuracy of pedestrian re-identification is further improved.