The use of multi-source, multi-objective statistical information fusion is utilized in this paper to identify garment design targets, and the correlation is judged based on garment data. Some parameters are pre-set to generate linear garment descriptions for multi-source, multi-objective information fusion and find the adaptability of ethnic elements under random coefficients using element information relationships. The clothing width vector of the radial basis function is calculated by clustering ethnic culture element data using centroid information. The results show that the innovation rate of ethnic elements of Miao batik reaches the highest 0.57, the average adaptation value in the garment design is increased by 4.92%, and the ethnic-cultural elements are embodied 280 times, which indicates that the proposed method can promote the traditional excellent culture.