Autonomous Sign Reading for Semantic Mapping on the Stanford AI Robot

<a href="https://embed.stanford.edu/iframe/?url=https%3A%2F%2Fpurl.stanford.edu%2Fxn156wf1698" class="su-underline">Show Content</a>

Abstract/Contents

Abstract: This thesis considers the problem of automatically collecting semantic labels during robotic mapping by extending the mapping system to include text detection and recognition modules.In particular, it describes a system by which a SLAM-generated map of an office environment can be annotated with text labels such as room numbers and the names of office occupants. These labels are acquired automatically from signs posted on walls throughout a building using a feed-forward system that collects a set of building images; uses a trained classifier to detect candidate text regions on those images; runs optical character recognition on those regions to extract the textual content; and post-processes the text string set with alignment information before placing each string at its original location on the map. Such a system faces difficulties using current text recognition techniques, as standard approaches to optical character recognition fail when faced with text from natural images as opposed to document text. Despite these difficulties, our system provides a series of additions to the typical mapping pipeline which provide practically useful results. In fact, our text detection and recognition system, combined with other ingredients, allow the robot to generate an annotated map from which it can recognize named locations specified by a user 84% of the time.

Type of resource	text
Date created	2011-05

Author	Case, Carl
Advisor	Ng, Andrew Y.
Department	Stanford University. Department of Computer Science.

Subject	Robots > Control systems
Subject	Autonomous robots
Subject	Semantic networks > Information theory
Genre	Thesis

Location	https://purl.stanford.edu/xn156wf1698

Use and reproduction: User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
License: This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).

Preferred Citation: Case, Carl (2011). Autonomous Sign Reading for Semantic Mapping on the Stanford AI Robot. Stanford Digital Repository. Available at http://purl.stanford.edu/xn156wf1698

Undergraduate Theses, School of Engineering

Loading usage metrics...