Corner Feature Extraction: Techniques for Landmark Based Navigation Systems

Abstract

In this paper we discussed the results of an EKF SLAM using real data logged and computed offline. One of the most important parts of the SLAM process is to accurately map the environment the robot is exploring and localize in it. To achieve this however, is depended on the precise acquirement of features extracted from the external sensor. We looked at corner detection methods and we proposed an improved version of the method discussed in section 2.1.1. It transpired that methods found in the literature suffer from high computational cost. Additionally, there are susceptible to mapping ‘ghost corners’ because of underlying techniques, which allows many computations to pass as corners. This has a major implication on the solution of SLAM; it can lead to corrupted map and increase computational cost. This is because EKF-SLAM’s computational complexity is quadratic the number of landmarks in the map, this increased computational burden can preclude real-