Home // International Journal On Advances in Software, volume 10, numbers 3 and 4, 2017 // View article

A New Approach for Vehicles Detection in Low Altitude Aerial Images based on Edge Density

Authors:
Antonio J. R. Neves
Manuel Camarneiro
Lucas Cozinheiro

Keywords: Drones; Image processing; object detection

Abstract:
Drones and Computer Vision are both important research topics nowadays. Combining these two technologies enables having different perspectives of a scene and capturing them with digital cameras onboard drones. Incorporating onboard processing of the acquired images might provide extremely useful information from different scenarios, allowing the drones to perform autonomously with applications in several contexts. In this paper, we present algorithms to process images captured by drones over parking lots in order to detect parked vehicles and further estimate occupancy rates or cars parked in a wrong place. As far as we know, low altitude image processing is still an open problem in the computer vision community. Experimental tests with the developed algorithms were performed in different parking lots across the University of Aveiro Campus under different lighting conditions to check their accuracy and processing requirements. Detailed experimental results are presented in this paper and an image database was produced in order to allow future experiments by other researchers. The obtained results presented in this paper demonstrate the effectiveness of the proposed approaches.

Pages: 490 to 498

Copyright: Copyright (c) to authors, 2017. Used with permission.

Publication date: December 31, 2017

Published in: journal

ISSN: 1942-2628