A gesture recognition system based on capacitive sensor FDC2214 is studied. In order to improve the accuracy of gesture recognition

the capacitance induction test method based on hand distribution is designed and the data acquisition strategy of multi-channel sensor is proposed. In the design of gesture training and recognition algorithm

bubble sorting method is proposed to optimize acquired gesture data. Using the method of the Delta difference operation of the new and old data and the absolute value accumulating

the gesture with the smallest weight is obtained. Gesture recognition device is tested for hundreds of times in the guessing and rowing games

respectively. The recognition success rate of the guessing games is 100 %

the recognition success rate of the rowing games is over 99. 2 %

and the recognition time of a single gesture is not more than 427. 9 ms

which achieves better real-time gesture recognition effect.