Collection of road traffic information from satellite images and digital map

Abstract

ABSTRACT There have been many reports on the analysis of the Earth’s surface by remote sensing. The purpose of this study is to analyze traffic information, and we have been studying methods of collecting traffic information by remote sensing. To collect traffic information, sensors installed on the roadside are frequently used. However, methods using sensors only collect information around the positions of the sensors. In th is study, we attempt to solve this problem by using satellite images, which have recently become increasingly available. We propose a method of collecting traffic information over a large area using satellite images as well as three-dimensiona l digital maps. We assess traffic conditions by computing the number of edges of vehicles per road section as follows. Fi rst, the edges of vehicles ar e detected in satellite images. During this processing, three-dimensional digital maps are used to increase the accuracy of vehicle edge detection. The number of vehicles per road section, which is computed from the number of edges of vehicles, is computed and referred to as the vehicle density. Traffic conditions can be assessed from the vehicle density and are considered useful for collecting information on traffic congestion. In this study, we experimentally confirm that congested roads can be extracted from satellite images by our method. Keywords: satellite image, three-dimensional digital map, road conditions