الفهرس 
FEATURE EXTRACTION AND DIGITAL IMAGE PROCESSINGOnly 14 pages are availabe for public view
 |  | from | 32 |  |  |
| | | from | 32 | | |
Abstract
Intoday’s world, mobility is considered a very important factor in every field. This has caused the
number of vehicles to increase on a daily basis. However, the capacity of road networks is limited
and the route planning is not optimized.
Naturally, this has led to the well known problems of urban traffic congestion. And there are two
ways to deal with these problems; either by optimizing the road networks’ capacity or by optimizing
the road networks’ efficiency through the Vehicle Navigation System (VNS).
VNS is expected to reduce drivers’ stress all over the road network. In addition, the ’Urban
Emergency’ response and the ’Disaster Management Systems’ require quick and reliable vehicle
navigation systems, in order to improve road safety.
Any efficient VNS should have three main modules: the first is the positioning module used for
tracking the vehicle’s position; the second is the communication module that is used for updating
the vehicle with the necessary information on call. The third one is the location module which is
designed to convert the determined position into useful information (location, direction ...). This
module has different processes; the most applicable ones are map matching and routing functions.
The main target is to provide the functions of the third module, where the research concentrates on
the following three main functions:
(1)Road extraction as a step of data preparation where the center line of the roads is extracted
”road segments”; and the seed extraction process as the points of intersection between road
segments are defined in order to get road theme invector format;
(2)0n-Road map matching which is the process of snapping the vehicle position to the extracted road
segments in order to display the snap vehicle position on the proper road segment, and
(3)0n-Time routing, that is considered one of the most important services offered by navigation
system which grants optimum path between specified source point to a specified destination node and
also, route guidance to guide the driver by a series of visual instructions.
Like navigation tools, the navigation system and the different sensors applied in it are outlined.
The related techniques as the general principles of graph theory and algorithmic techniques are
presented. Moreover, map matching techniques are highlighted. Also, the basic theory and the
evolution of digital image processing and road extraction are described. Adequate algorithms are
derived and implemented with developed software in order to achieve our target.
In order to evaluate the behavior and performance of the proposed algorithm samples of data are
applied with simulation tests to each sub-module. Generally, vehicle navigation system is a real
time application that requires high speed and performance of the sub-modules, so, the speed and
flexibility of the applied packages are put into consideration.