Fuzzy Inference Integrated 3-D Measuring System With LED Displacement Sensor and Vision System

Abstract

This article deals with a 3-D measurement system applied to a curved metal surface carving system and a sensor integration method based upon fuzzy inference. The measurement system consists of two different sensors. One is a LED displacement sensor, while the other is a vision system. The LED displacement sensor's spotlight is used as a part of the vision system based upon the active stereo sensing method. In addition, the LED displacement sensor's outputs are for calibrating camera parameters. Therefore, we can calibrate the camera parameters easily. Then, we use neural networks to compensate the output of the image processing for some errors, such as camera parameter's error and lens distortion. By utilizing neural networks, we can use a vision system as accurately as possible. We use a sensor integration method based upon the fuzzy set theory. Fuzzy inference's input consists of information on the change in the sensor output and the position change of the sensor system, together with an environmental data of measurement. For this integration system, we can use the sensory system accurately. The proposed system is shown to be effective through extensive experiments.