Multi-Sensor Data Fusion for Autonomous Vehicle Navigation and Localization through Precise Map

Abstract

Inexpensive implementation of localization and environment mapping are critical issues for urban autonomous driving. We present a practical and low-cost navigation architecture to fuse different data from vehicle onboard sensors and estimate the vehicle state when individual observations such as GPS are noisy. We are trying to compensate the GPS errors by data fusion from different sensors in a probabilistic way and a particle filter with joint observations model has been proposed. We have evaluated the feasibility of proposed localization and navigation architecture for fully autonomous driving by doing many experiments in our campus including up and down slopes.