Error backpropagation neural calibration and Kalman filter position estimation for touch panels

This work develops a methodology and technique for calibration and dynamic touching position estimation of touch panels using error backpropagation neural networks (EBPNN) and Kalman filter. A neural-based calibration method is presented to determine the nonlinear mapping relationships of the measured and known touch points, and then calibrate their positions in a real-time manner. In order to obtain position estimation of fast moving points in the drawing mode, a Kalman filtering scheme is proposed to achieve a satisfactory precision.