With the booming development of the cultural and creative industry (CCI), the demand for the design of cultural and creative products (CCP) is growing. However, traditional design models have problems such as low efficiency and limited creativity, making it difficult to meet the diverse needs of the market. Aiming at the complex needs of image style transfer for CCP, this paper designs an image style transfer algorithm that integrates semantic features. The algorithm uses a pre-trained deep convolutional neural network to extract the content and style features of the image, introduces a semantic segmentation model to generate a semantic feature map, and constructs a semantic loss function to force the network to retain the semantic integrity of cultural elements during the style transfer process. At the same time, the algorithm integrates multi-scale feature fusion and spatial attention mechanism to improve the algorithm's ability to capture image details. Experimental results show that the platform can quickly generate cultural and creative product images of diverse styles. In terms of dynamic loss weight adjustment, the platform model improves the realism of style transfer by about 40% while maintaining content integrity. The actual conversion rate of cultural and creative design cases is as high as 68%, and the design efficiency is significantly improved, providing an effective solution for the intelligent development of the CCI.