Towards a human airbag system using /spl mu/IMU with SVM training for falling-motion recognition

A micro inertial measurement unit (muIMU) which is based on MEMS accelerometers and gyro sensors is developed for real-time recognition of human body motions, specifically falling-down motions caused by slippage. A muIMU measures three-dimensional angular rates and accelerations. With an integrated microcontroller, the overall size of our muIMU is less than 26 mm*20 mm*20 mm. We present our progress on using this muIMU based on support vector machine (SVM) training to recognize falling-motions. The digital sample rate of the micro controller is 200 Hz which ensures rapid reaction to short falling time and also gives a sufficient data information for SVM recognition. Experimental results show that our system can achieve a lateral falling-motion recognition rate of 100% using selected eigenvector sets generated from 200 experimental sets. Our goal is to implement this system to a human airbag system designed to protect hip fractures of the elderly