Trajectory Reconstruction Of Writing Brush Movement Based On Inertial Sensor

Calligraphy is a Chinese character writing art formed by black and white.It is the first of Chinese arts with paintings.It is a symbolic representation of the traditional culture of the Chinese nation and a treasure of the East.Combining electronic technology with digitalization of calligraphy creation is of great significance for in-depth understanding of calligraphy and inheritance calligraphy.The digitalization of calligraphy creation includes the digitization of brush,ink,thesis,and cymbals,among which the digitization of the brush includes the digitization of the brush head and brush tail.In order to study the relationship between the movement of the brush pen and the brush strokes and achieve the goal of digitizing the brush,this thesis reconstructs the trajectory of the pen based on the inertial data collected by the inertial sensor.This thesis designed and manufactured electronic brush hardware and corresponding inertial data acquisition software.The electronic writing brush fixes the inertial measurement module on the pen and records the inertia data corresponding to the movement of the brush pen.Inertia data is transmitted through the Bluetooth module,and recorded by the acquisition software in the upper computer.Using inertial data and the principle of strapdown inertial navigation to reconstructs the trajectory of the brush pen in the upper computer.The inertial data includes triaxial acceleration,angular velocity,and magnetometer data.Firstly,the inertial data was preprocessed,including removing abnormal data,ellipsoid fitting,time series modeling and other methods,which reduced the measurement data error.Then,a trajectory reconstruction scheme using one inertial sensor was designed.The acceleration,angular velocity,and magnetometer nine-axis data were combined to perform the attitude calculation.The zero-speed detection and Kalman filtering were combined to eliminate the cumulative error.The trajectory reconstruct experiment was simulated using the artificial data and the acquired data.Then,in order to improve the accuracy of trajectory calculation,a trajectory reconstruction scheme using two inertial sensors is designed,based on the two inertial sensor installation structures in the electronic brush designed,the position error is calculated by using the distance constraint relationship to reduce the accumulated error.At the same time,the constraint relationship is used to improve the elimination of abnormal data and reduce the data measurement error.The experimental results show that the trajectory reconstruction accuracy of the two-sensor trajectory reconstruction scheme is higher than the single sensor scheme.For noise-less artificial data,both schemes can effectively eliminate accumulated errors and obtain accurate trajectories,verifying the correctness of the method.For the noisy artificial data and the collected inertial data,the accuracy of the two sensors is higher than that of a sensor.The trajectory result still has about 10% error,which shows that the design method is sensitive to the noise of the measurement data.The analysis shows that the civil-level inertial sensors used in this thesis have a relatively large amount of noise.The stability of the trajectory reconstruction of the pen is poor.