Abstract: Limitations are found in the personalization of patterns in traditional footwear design, and there is a lack of stylized adjustment scheme for generating new texture patterns. Style transfer was employed to achieve the personalized design of footwear patterns, and a method of personalized style transfer for patterns with footwear region control was proposed to address the issue that existing style transfer models perform poorly on footwear products. The main body of the footwear and the background were extracted through an image segmentation algorithm that combines a fully convolutional network and a conditional random field. Binarization processing was then carried out to obtain a mask image for subsequent spatial control. Subsequently, a style transfer model based on Visual Geometry Group 19 was combined with the mask to realize the style transfer from the style image to the target regions of the footwear. The experimental results show that, compared with style transfer methods such as Adaptive Instance Normalization, this method achieved better results in the generated style transfer effect images of footwear products. It solved problems such as the unclear boundaries between the footwear products and the background after style transfer, and effectively improved the application effect in the personalized design of footwear patterns.