Abstract After a long period of development and evolution, Chinese painting has formed a set of regular artistic language, which highlights the national characteristics of Chinese painting and is an important factor distinguishing it from other works of art. As the main characteristic of Chinese painting, this paper analyzes the traditional techniques of Chinese painting according to the three characteristics of Chinese painting: line, ink, and color, and analyzes the language of programed techniques. Then the method of digital reproduction of traditional techniques of Chinese painting is proposed and elaborated with the example of “The Western Garden of Elegant Gathering”. Subsequently, for the complexity of hand movement recognition of Chinese painting techniques, this paper proposes a sample splitting and re-fusion algorithm RCF based on sEMG to optimize the hand movement signals and construct a hand movement recognition model based on surface EMG signals. Through empirical analysis, the SVM classifier used in this paper has an average recognition correctness rate of 98%, which is 6% and 3% higher than BP and LDA, respectively. Meanwhile, the recognition accuracy of the hand movements of the eight techniques is all above 95%, with an average recognition accuracy of 98.5%. The mean value of all five indicators of the experience evaluation of Chinese painting techniques reproduction and innovation exceeded 4 points, and the overall average score was 4.542, indicating that the experiencers were very satisfied with the experience of Chinese painting techniques reproduction to innovation in this paper. The research of this paper provides theoretical and technical guidance for the reproduction and innovation of traditional Chinese painting techniques, as well as ideas for the organic combination of digital technology and traditional Chinese paintings.