A movement recognition method based on a surface electromyographic signal comprises: obtaining surface electromyographic signals from multiple channels (101, 301)； determining an effective surface electromyographic signal on the basis of the surface electromyographic signals of the multiple channels (102, 302)； determining the frequency of the effective surface electromyographic signal (103, 303)； determining body movements corresponding to the surface electromyographic signals of multiple channels on the basis of the frequency of the effective surface electromyographic signal (104). Since the frequency of the surface electromyographic signal is independent of features such as the signal intensity and the like, the method can greatly increase the accuracy of the movement recognition based on the surface electromyographic signal. Moreover, since the frequency is taken as the recognition feature, users do not have to make excessively large movements so as to enjoy a better user experience. In addition, the movement recognition method based on the surface electromyographic signal can further comprise: extracting the amplitude feature of the effective surface electromyographic signal (304)； determining body movements corresponding to the surface electromyographic signals of multiple channels on the basis of the amplitude feature of the effective surface electromyographic signal and the frequency of the effective surface electromyographic signal (305). The method not only can increase the accuracy of surface electromyographic signal recognition, but also can increase the number of body movement recognition types.La présente invention concerne un procédé de reconnaissance de mouvement basé sur un signal électromyographique de surface, qui comprend : l'obtention à partir de multiples canaux (101, 301) de signaux électromyographiques de surface ； la détermination d'un signal électromyographique de surface efficace sur la base des signaux électromyographiques des canaux mult
