Research And Design Of Ppt Control System Based On Dynamic Gesture Recognition

With the popularity of mobile smart devices,wearable electronic devices have been part of modern life,and people require more strictly for human-machine interaction experience of electronic devices.The MEMS acceleration sensor based gesture recognition technology which has great value in research and economic benefit is a hot topic in the research of the new generation of human-machine interaction.Based on the MEMS accelerometer,this study developed a PPT control system that can help to change slides by recognizing simple gestures.The main tasks of this study include:1)After analyzing the activity characteristics of human gestures,six commonly used dynamic gestures are defined based on the principles of operability and identifiability,and these six gestures respectively correspond to the four common slides operations.A calibration method is proposed for the deterministic error of MEMS accelerometers,and the Allan variance analysis and software compensation filtering are performed for the random error of the MEMS accelerometer to ensure the accuracy of the algorithm.2)The identification method based on feature extraction is determined by comparing and analyzing four current gesture recognition algorithms.After pre-processing the acquired data,which mainly skip point removing and moving average filtering,the key point information of the gesture data is intercepted by setting a threshold value,and then the applied in normalized processing.Finally,feature extraction is performed.Four feature quantities,including gesture length,peak number,maximum axis of gesture energy,and peak or trough,were defined to distinguish the gestures,and individual differences are largely avoided.3)The practical requirements of the control system in the PPT presentation are analyzed.The commonly used MMA7361 triaxial accelerometer in the laboratory was selected as the acceleration data acquisition source.Microprocessor MSP430F149 was used as the core of data processing,and NRF24L01 was the wireless communication unit.On this basis,the hardware and software system were designed in detail.The result which showed the good recognition was verified by virtual oscilloscope.Compared with traditional handheld devices,the designed control system in this study has high sensitivity and high recognition accuracy.It is easy to carry and spends short time to respond.Therefore,it is confirmed to have value for applying.