The digital virtual design model in the field of cultural and creative product design has attracted much attention of researchers due to its high adaptability. The traditional cultural and creative product design method has poor color feature interpolation effect, so algorithm improvement is urgently needed. Based on the theory of partial differential color feature extraction, this article proposes an improved design method for segmenting digital virtual cultural and creative products and introduces a color feature item on the basis of the original model. The improved partial differential algorithm can not only solve the re-initialization problem in the process of feature evolution, but also adopt a more flexible initialization method (the level set partial differential function can be loosely defined as the signed distance partial differential function). The simulation results show that the partial differential color feature extraction algorithm has a good performance in determining the target area of the product, the target segmentation rate reaches 91.7%, and the calculation speed of the algorithm is increased to 0.216s, which effectively improves the color feature interpolation effect of cultural and creative products by bringing results closer to the digital virtual trend.