Robot Localization in RGB-D Images Using PCA and CNN

Abstract

Human beings have always used the resources at their disposal as tools to help carry out tasks in the most effective, fast and safe way. As technology advances, these tools are being used in increasingly complex machinery capable of performing complicated jobs precisely and often better than a human could have in this paper we aim at performing a robust robot localization by reducing the 3d images into 2d images using PCA algorithm and using CNN for feature extraction and classification of the images.