The traditional clothing patterns of the She ethnic group are unique in the Chinese minority pattern art because of its rich decoration. This paper utilizes Convolutional Neural Networks (CNN) to construct the YOLOv8 algorithm model for intelligent classification and recognition of the patterns, compares it with the VGG16 model, and finally conducts a comparative analysis of the experimental results. The findings of this paper reveal that the YOLOv8 model outperforms the VGG16 model in most classification tasks, with an average accuracy of 59.13%. Using CNN for classification and recognition can reduce the complexity while enhancing the speed and accuracy of classification, providing a new approach for the protection and inheritance of traditional patterns of ethnic minorities.