An Improved Scan Matching Method Based on Laser Reflection Intensity

Abstract

High-precision localization performance of mobile robots in featureless environments plays an important role in the field of navigation. This paper describes an effective localization solution using a landmark whose surface is covered with two different materials resulting in different laser reflection intensities. Based on the landmark, a feature extraction algorithm is proposed and an improved scan matching method for mobile robot localization in featureless environments is described. Unlike traditional scan matching methods which only use geometric features, the proposed method mixes geometric information extracted from environment and laser reflection intensity extracted from the landmark to achieve accurate pose estimation. Experiments show that our method has higher precision than previous methods.