Abstract
It is important technology for an autonomous mobile robot to generate the environmental map and to estimate its position and posture. In recent years, the environmental measurement using LRF is increasing, because measurement by the LRF is a high-speed and its data is very accurate. However, since LRF acquire only 2D data at once, in order to obtain 3D data, it is necessary to rotate a sensor. Shape of the 3D data acquired by rotating LRF becomes slit. When these data are connected by ICP algorithm, an error occurs by the shape-dependence property of it. In this paper, we discovered that the parameter has an error is different in the high area and the low area of the 3D data. Then, we propose a method for improving the accuracy of mapping by integrating the parameters without error, after matching in each of the high area and low area of data. And we show the utility of the method by experiment.
