Biorobotic Vision Laboratory

Mandyam V. Srinivasan
Javaan S. Chahl
Matt A. Garratt
Akiko Mizutani
Dean Soccol
Geno Ewyk

Over the past 10 years, research in our Insect Vision, Navigation and Cognition Laboratory and elsewhere has revealed a number of principles by which insects see, control flight, and navigate. The broad aim of the Biorobotics Vision Laboratory is to use these principles to devise novel algorithms for machine vision, and novel kinds of autonomous, visually-guided robots. In particular a number of algorithms have been devised to use optic flow to measure image motion.

Recent work has focussed on developing autopilots for unpiloted airborne vehicles (UAVs) based on vision.

We have designed and built


A visually mediated hover controller that can hover over an arbitrary set of landmarks.

We have also developed and patented a Panoramic Imaging System [14,20]. The system comprises a standard video camera viewing a specially shaped reflective surface. The surface has the property that a given change in the angular elevation of view in the external environment maps to a constant radial displacement in the camera's image. This system has potential applications in surveillance systems as well as in visual guidance and control of autonomous vehicles.

A corridor-following robot [18]Download an MPEG movie of the robot in a corridor.


A terrain following autopilot for an autonomous helicopter, based on optic flow

A robot that measures visual image deformation to estimate its own motion in the environment [9]Download an MPEG movie of the robot in action.

An obstacle-avoiding robot [6]

The laboratory has devised

Selected publications from the Biorobotic Vision Laboratory

----------------------------------------------------------------------

see B-EYE for a bee's-eye view of the world

Go Back to CVS Page