Research And Application Of Gesture Recognition Based On Inertial Sensor Of Wearable Device

Along with the continuous promotion of wearable devices,more and more people are starting to use wearable devices such as smart watches and smart wristbands,and more and more technologies are being applied to such smart devices.Wearable devices have many functions such as motion monitoring,sleep monitoring,notification reminders,time display,authentication,payment,and gesture recognition.Gesture recognition is an important interactive method in human-computer interaction.Through gesture recognition,it can replace the original user interface and graphical user interface to perform faster and more convenient operation.At present,the data collection form of gesture recognition mainly uses two kinds,one is based on image video data,and the other is based on inertial sensor data,and the wearable device naturally uses the data of inertial sensors on the device,based on the inertial sensors.The axis acceleration data is related to hand gesture recognition.Gesture recognition process is divided into the collection of raw data,the pretreatment of the original data,and finally gesture recognition.In the process of preprocessing,there are many details that may have a significant impact on the final recognition effect.Therefore,through the comparison of details such as noise reduction processing and effective data interception,it is further analyzed how to make the processed identification rate higher.The dynamic time warping(DTW)was improved,and the recognition method with higher accuracy was achieved by combining with K's nearest neighbor classification algorithm.By setting the optimal K value,the recognition accuracy rate was the highest.The filtered standard template library verifies its universality.