Lane Departure Warning and Object Detection Through Sensor Fusion of Cellphone Data

Examensarbete för masterexamen

Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.12380/223154
Download file(s):
File Description SizeFormat 
223154.pdfFulltext9.95 MBAdobe PDFView/Open
Type: Examensarbete för masterexamen
Master Thesis
Title: Lane Departure Warning and Object Detection Through Sensor Fusion of Cellphone Data
Authors: Eriksson, Jesper
Landberg, Jonas
Abstract: This master thesis focus on active safety for the automotive industry. The aim is to test an inexpensive implementation of some common functions realized using a cellphone to gather data. A Matlab Simulink model is developed for the purpose, and then the agility of the model is tested by generating c code and running it on a single board computer. A robust lane detection algorithm is developed by using Hough lines. To better cope with curves in the road, the Hough lines are combined with a parabolic second degree fitting. The Hough lines are also used for a Lane Departure Warning system. Using edge filtering and connected component labeling an obstacle warning is implemented. Overall the model works well and is fast enough to meet the real time requirements when run on a computer. On the Raspberry Pi 2 chosen as the single board computer the processing is unfortunately not quite fast enough for high speed driving. However when the object detection is removed the Raspberry Pi 2 meets the real time requirements as well.
Keywords: Farkostteknik;Hållbar utveckling;Innovation och entreprenörskap (nyttiggörande);Transport;Vehicle Engineering;Sustainable Development;Innovation & Entrepreneurship;Transport
Issue Date: 2015
Publisher: Chalmers tekniska högskola / Institutionen för tillämpad mekanik
Chalmers University of Technology / Department of Applied Mechanics
Series/Report no.: Diploma work - Department of Applied Mechanics, Chalmers University of Technology, Göteborg, Sweden : 2015:03
URI: https://hdl.handle.net/20.500.12380/223154
Collection:Examensarbeten för masterexamen // Master Theses



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.