Chuannan Miao weaving technique is a precious cultural heritage of the Chinese nation, carrying thousands of years of cultural deposits of the Miao people, and its traditional display method has problems such as poor interactivity and weak experience. For this reason, this paper proposes a digital display system based on the fusion of virtual reality and augmented reality technology for the weaving techniques of the Miao people in Chuannan. Methodologically, the study adopts the improved G-SIFT algorithm to replace the traditional Gaussian function, uses the guided filter function for feature extraction, and constructs a 4-layer architecture system based on J2EE Web technology. Experimentally selected three types of images of embroidery, batik and brocade are verified, and the results show that: in terms of feature extraction matching efficiency, this paper’s algorithm outperforms the traditional SIFT, SURF, ORB, and AKAZE algorithms on the three types of images; in terms of the evaluation indexes, the GSIFT algorithm shows significant improvement in the information entropy, the mean squared error, and the generalized image quality index; in the system performance test, the The average response time for concurrent login of 5000 users is 2.149s, the response success rate is 100%, the CPU utilization rate is lower than 50%, and the memory utilization rate is lower than 60%. The conclusion shows that the VR/AR fusion technology based on G-SIFT algorithm can effectively extract the characteristics of weaving techniques, and the constructed digital display system has good performance, which provides a new technical path for the inheritance and protection of the weaving techniques of Sichuan Miao.
