Calibration of MEMS accelerometer based on plane optical tracking technique and measurements

Abstract

A MAG-µIMU which is based on MEMS accelerometers, gyroscopes and magnetometers has been developed for real-time estimation of human hand motions. Appropriate filtering, transformation and sensor fusion techniques are combined in the Ubiquitous Digital Writing Instrument (UDWI) to record the handwriting on any surface. However, because of the sensors' intrinsic biases and random noise such as circuit thermal noise, a calibration system that provides good reference measurement parameters must be used to compare the output of the MAG-µIMU sensors. In our previous work, a novel idea to calibrate three-dimensional linear accelerations, angular velocities and space attitude through optical tracking techniques had been proposed and proved. In this paper, we will introduce the latest progress in calibrating the accelerometers. In the experiments, the entire English alphabet has been written down by UDWI and these hand-writing motions are recorded by a high speed camera with the same sampling rate as the microcontroller. Then by dividing each image sequence into single frame, the blocking matching algorithm can help us get position information, based on which another group of velocity and acceleration are able to be obtained to compare with the sensor output to calibrate. From the comparison, we can find out how the sensor drifts during the integration. In addition, the consistency property of the sensors is also tested by analyzing a single letter written multiple times. At last, the resolution problem of the calculation of velocity and acceleration by the means of optical tracking measurement is discussed.