Badminton is a sport that is very popular, and it has many benefits to the human body. At present the scope of the research on badminton movement using inertial senors is mainly focused on the relationship between the dynamic speed and speed of the racket, or the dynamic analysis of the swing arm of the badminton players.and There are currently few studies on badminton strokes identification, but a lot of study on the relationship between dynamic speed and racket speed or the dynamic analysis of the swing arm of the badminton players. In this research, they offer an initial online identification system for badminton strokes, and they show how to collect badminton data using an inertial sensor attached to the bottom of a racket. Extracting data on badminton strokes from video sequences is done using a threshold-based detection method and sliding window methodology. For the classification phase, they retrieved a total of twelve characteristics to make up the feature space, and they employ random forest (RF) to identify badminton shots. Experiments with a total of 12 participants show that the proposed algorithm is both successful and useful.