Simultaneous Localization and Mapping (SLAM) : towards an autonomous search and rescue aiding drone

Heukels, F.R (2015) Simultaneous Localization and Mapping (SLAM) : towards an autonomous search and rescue aiding drone.

[img]
Preview
PDF
6MB
Abstract:When having to find survivors in a heavily damaged building it is saver to send a drone that generates a 3D map before sending in emergence personnel. The generated map can then be used to locate the survivors and rescue them. This thesis proposes a light weight implementation of a 3D mapping algorithm that is designed to be run onboard of a drone. The map is generated by stitching together multiple images from a RGBD (RGB and depth image) camera by looking at the movement of the drone between two consecutive frames. This is accomplished by calculating an initial transformation by using ORB features and then optimizing this transformation using the point to projection scan matching method. To be able to test the implementation a dataset is recorded by moving the camera through a hallway. The implementation is able to stitch together all the images into a single 3D model with an average frame rate of 17 fps on a HP 8570w. There is however an ever-increasing error that can be reduced by using an Inertial Measurement Unit (IMU) or by implementing a loop closing algorithm or by using graph optimization techniques.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Embedded Systems MSc (60331)
Link to this item:http://purl.utwente.nl/essays/66615
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page