Method for hiding multiple high-resolution color images based on Swinin-Unet architecture

The invention relates to a method for hiding multiple high-resolution color images based on a Swinin-Unet architecture, and the method comprises the steps: S1, collecting an image data set to construct a training data set, carrying out the preprocessing and data enhancement to construct an enhanced data set, and carrying out the cascading of a plurality of secret images and a single carrier image in each iteration along a channel dimension; s2, constructing a coding network based on Swinin-Unet, calculating self-attention for the activation graphs of different scales to perform hiding operation of the secret image, and outputting a coded image; s3, constructing a decoding network based on Swinin-Unet, calculating self-attention for the activation graphs of different scales to decode the secret image, and outputting a decoded decrypted image; and S4, carrying out iterative training, updating model parameters by optimizing combination loss, continuously storing an optimal model according to verification coding