Development and implementation of a paint brush algorithm for computer art

Abstract

The purpose of this research is to develope an electonic paint brush which can be used as an input device for computer art. To express various effect of brush touch, the 3-dimensional parameters such as elevation and azimuth are to be known. To detect the 3-dimensional movement, two ultrasonic sensors(transmitter) are attached to the paint brush. Two transmitters emit ultrasonic pulses alternately while the paint brush moves on the canvas. The four receivers located at the four edges of the canvas receive the pulses. From the time difference between transmitted and received pulses, the 3-dimensional state of the paint brush can be obtained. In this thesis, an algorithm, hardware, software have been developed to implement a paint brush system. The hardware is a PC slot board that sends two transmitters clock burst periodically for the transmitters to generate ultrasonic pulses, and measures the time difference between transmission and reception. Four receivers at the canvas edges receive the ultrasonic wave pulses which are converted subsequently to electrial signal by the sensors. By passing an amplifier and a comparator and a pulse shaper, the sensor output signal is transformed to a pulse whose width is equal to that of the signal. The measured time difference is sent to the algorithm part. Here the extended kalman filter algorithm is used to estimate the position of paint brush. The painting software draws the position of the brush on the screen. Users can select the color and thickness of the brush with mouse.