A study of angular estimation of the mobile robot by using the sun azimuth angle based on omnidirectional image

Abstract

The reliable moving direction of the mobile robot helps stable navigation as well as robust self-localization. Since the sun is one of visible dominant landmarks for outdoor navigation, in this paper, we describe a new estimation algorithm for the mobile robot direction based on the sun azimuth angle from omnidirectional image. The proposed approach can be applicable outdoor navigated by using only omnidirectional camera without other extra sensors. Results are provided which show that our technique can estimate the relative moving direction of the mobile robot with good accuracy that is comparing to the sun azimuth angle.