Unsupervised pedestrian re-identification method, device and equipment and readable storage medium

The invention provides an unsupervised pedestrian re-identification method, device and equipment and a readable storage medium. The method comprises the following steps: clustering pedestrian image samples, and distributing a false label for each pedestrian image sample based on a clustering result; constructing a category feature memory and a feature memory; obtaining K * P pedestrian image samples according to the false labels; inputting the preprocessed K * P pedestrian image samples into a deep learning model to obtain second pedestrian features; optimizing a deep learning model based on the category feature memory, the feature memory and the second pedestrian feature; and when the number of optimization times reaches a preset number of times, obtaining pedestrian features corresponding to the target pedestrian image by using a finally obtained deep learning model, and matching the pedestrian features corresponding to the target pedestrian image with pedestrian features in a database to obtain a pedestria