A Low-Cost Mirror-Based Active Perception System for Effective Collision Free Underwater Robotic Navigation
Abstract
This ongoing research work presents a servo actuated mirror-based design that allows a fixed front-view visual system mounted in an underwater vehicle to extend its field of view by controlling its gaze. We are interested in the autonomous underwater exploration of coral reefs. This type of exploration must involve a cautious and collision-free navigation to avoid damaging the marine ecosystem. Generally, vision systems of underwater vehicles are carefully isolated with mechanical seals to prevent the water from entering. However, this fact causes a strictly dependence between the angle of view of the camera and the pose of the vehicle. Furthermore, the addition of a system to control camera orientation may result in a significantly reduction of useful load capacity and the movement of the vision system could carry undesirable trusting effects, especially at higher speeds. Our design of servo actuated mirror system changes the angle of view of the camera in two degrees of freedom: pan and tilt, and reaches viewing angles from the sides, bottom top and even rear views of the robot, thus enabling a more effective navigation with obstacle avoidance.
Cite
Text
Cortés-Pérez and Torres-Méndez. "A Low-Cost Mirror-Based Active Perception System for Effective Collision Free Underwater Robotic Navigation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2016. doi:10.1109/CVPRW.2016.15Markdown
[Cortés-Pérez and Torres-Méndez. "A Low-Cost Mirror-Based Active Perception System for Effective Collision Free Underwater Robotic Navigation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2016.](https://mlanthology.org/cvprw/2016/cortesperez2016cvprw-lowcost/) doi:10.1109/CVPRW.2016.15BibTeX
@inproceedings{cortesperez2016cvprw-lowcost,
title = {{A Low-Cost Mirror-Based Active Perception System for Effective Collision Free Underwater Robotic Navigation}},
author = {Cortés-Pérez, Noel and Torres-Méndez, Luz Abril},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2016},
pages = {61-68},
doi = {10.1109/CVPRW.2016.15},
url = {https://mlanthology.org/cvprw/2016/cortesperez2016cvprw-lowcost/}
}