3D scene reconstruction and mapping with real time human detection for search and rescue robotics

With this work we aim to create a robust robotic and advanced computer vision software infrastructure aimed to enhance search and rescue algorithms in robotics and vision by introducing a new dimension to perception. For the scope of this project we will use the third dimension to generate detailed...

Full description

Saved in:
Bibliographic Details
Main Authors: Gautham, J.S., Sharma, Akash, Dhanalakshmi, Samiappan, Ramamoorthy, Kumar
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Computer vision
Disasters
Environment models
Image reconstruction
Multiple robots
Real time
Robotics
Search and rescue missions
Three dimensional models
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With this work we aim to create a robust robotic and advanced computer vision software infrastructure aimed to enhance search and rescue algorithms in robotics and vision by introducing a new dimension to perception. For the scope of this project we will use the third dimension to generate detailed 3D models of an environment and hence using it to detect survivors and segment human subjects from a chaotic disaster site in an efficient real time implementation. This also allows us to mark the position of human subjects in 3D space that can be transformed to a world frame. This world frame can then be shared with multiple robots that can localize the search and rescue subject humans in world coordinates instead of camera frame allowing for a 3D map of the disaster environment.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0100918