Bilkent University
Department of Computer Engineering

---

S E M I N A R

 

Using Shape Information From Natural Tree Landmarks for Improving Slam Performance

 

Bilal Turan
MSc Student Student
Computer Engineering Department
Bilkent University

Localization and mapping are crucial components for robotic autonomy. However, such robots must often function in remote, outdoor areas with no a-priori knowledge of the environment. Consequently, it becomes necessary for field robots to be able to construct their own maps based on exteroceptive sensor readings. To this end, visual sensing and mapping through naturally occurring landmarks have distinct advantages. With the availability of high bandwidth data provided by visual sensors, meaningful and uniquely identifiable objects can be detected, helping construct maps consisting of natural landmarks that are meaningful for human readers as well.

In this thesis, we focus on the use of trees in an outdoor environment as a suitable set of landmarks for Simultaneous Localization and Mapping (SLAM). Trees have a relatively simple, near vertical structure which makes them easily and consistently detectable. Bark surfaces associated with each tree are distinct enough to allow identification of each tree to address the data association problem for SLAM implementations. More importantly, however, the thickness of a tree can be accurately and consistently determined from different viewpoints. Our primary contribution is the use of this thickness information associated with the tree trunk as an additional sensory reading, allowing us to include the radius of the trunk as an additional piece of information associated with each tree in the map. To this end, we introduce a new sensor model that related the width of a tree landmark on the image plane to the radius of its trunk. We provide a mathematical formulation of this model, derive associated Jacobians and incorporate our sensor model into a working EKF SLAM implementation. We show through simulations that the use of this new sensory reading within a SLAM framework improves the accuracy of both map and the trajectory estimates without the need for adding additional sensor hardware other than a monocular camera.

 

DATE: 19 March, 2012, Monday @ 13:30
PLACE: EE314