Unsupervised pedestrian re-identification method based on shielded key area

The invention provides an unsupervised pedestrian re-identification method based on a shielded key area, which comprises the following steps of: preprocessing an unlabeled picture data set, and inputting the preprocessed picture data set into a network model; constructing a deep learning model, and using a space attention module to obtain a key area of the picture and shielding the key area; clustering the feature codes of the pictures to obtain pseudo labels of the pictures; constructing a loss function based on a difficult sample updating strategy and cluster updating; obtaining a trained network model according to the change condition of the loss function; and inputting a to-be-identified pedestrian picture and a to-be-identified pedestrian video into the trained network model, and outputting a pedestrian re-identification result. According to the method, the network model can be prevented from excessively paying attention to local features or global features of the image, and the generalization and robust