Data Registration and Integration of Three-Dimensional Sensors

Abstract

Based on a vertical scanning system with a line-structured light source, data registration of several 3-D sensors is performed by a calibration process using a linear partition method. Data integration is fulfilled by using a relatively simple but efficient technique called the average coordinates method. The principle and solution procedure of each method is deduced in detail. Two 3-D sensors, a round filament target and an object with certain edge length, are used to verify the theory, and experimental results show that the mean relative error is less than 1 %. With no requirement for solving nonlinear equations and other camera parameters, the method proposed is easy to adopt, runs fast, and still is high in precision, which meets the demands of 3-D measurement of a relatively big object