Automobile 360-degree look-around image rain removal method based on Transform network

The invention relates to an automobile 360-degree look-around image rain removal method, which comprises the following steps of: constructing a data set, and collecting 360-degree look-around images with information correlation and similar scenes as a network training data set; building a network model, wherein the network model comprises a convolutional neural network and a Transform network; extracting features of the network training data set input image through a convolutional neural network to obtain a feature map; paying attention to the feature elements in the feature map through a Transform network, and associating the feature elements with the most relevant feature elements; extracting features through a discriminator according to feature elements output by the Transform network, and multiplying the corresponding elements to generate an attention map, namely a rain-free image generated by the Transform network; training: inputting a rain-free image generated by the Transform network and a real rain-f