04-07-2012, 03:51 PM
Trac and Road Condition Monitoring System
[attachment=26716]
Introduction
The road network of any city is its lifeline and so monitoring this vital infrastructural resource
is important. The problem of trac jams and congestion are faced by most major cities of the
world. For managing trac, a city's administration should have both realtime and historical data
about the trac conditions prevailing on the road network. This information can be used for
quick reaction measures, such as, changing the timings of trac lights and advising commuters
to take alternative routes through public broadcasts. In the long run, however, this information
can be used to plan a better road network by identifying areas of frequent congestion and building
alternative routes.
Apart from managing trac on the roads, maintaining the road infrastructure in good condi-
tion is necessary. Municipalities generally have tight budgets. In developing countries like India,
funds are even scarce. Hence, what the authorities want to know is where, and to what extent is a
road damaged. This would enable them to take preventive measures before further damage occurs
or prioritize repair work based on the severity of damage. It is worth noting that damaged roads
with several potholes also lead to choking of trac and cause accidents.
Problem Statement
Our goal is to build a trac and road monitoring system for intelligent route planning, road usage
and maintenance that ful lls the constraints imposed by the Indian scenario. This system should
work under varied road conditions, chaotic, dense and unstructured trac and a large variety of
vehicles. It should be cost e ect, easy to deploy (no need to dig or build overhead structures) and
require minimal maintenance. We should avoid the need for specialized equipment. In order to
meet these somewhat con
icting requirements we are willing to be content with system that does
an approximate, aggregate trac analysis and near realtime reporting. We do not want a explicit
count or classi cation of vehicles but rather some information through which we can deduce the
state of trac on a road segment. Hence, we are willing to trade-o accuracy of reporting with ease
of deployment. We want to build a road monitoring system that is able to better quantify a road
anomaly. Thus, our e orts will be to try nd out ways to report severity, intensity or dimensions
of a pothole or a damaged road segment.
Magnetic Loops
Magnetic loop is a technology that has been used for vehicle detection and trac control for the
past few decades. These devices are installed inside each trac lane and act as counters, counting
vehicle passing over them. Some variants of the magnetic loop have been used to classify vehicles
as well. The loop is a continuous run of wire which is buried inside a trac lane. The ends of
the loop wire are connected through a loop extension cable to the vehicle detector. The detector
powers the loop causing a magnetic eld in the loop area. The magnetic
ux linked with the loop
changes whenever a metal object, such as a vehicle, moves over the loop. The detection scheme
of loops is based on this principle. The change in
ux is sensed by the detector which forces a
normally open relay to close. The relay remain closed until the vehicle leaves the loop.
Camera Based Systems
Camera based systems are able to detect, count and classify vehicles. These systems use video im-
age processors to identify vehicles and their trac
ow parameters by analyzing imagery supplied
by video cameras. Images supplied by cameras are digitized and then series of image processing al-
gorithms are applied on them. Information about vehicle passage, presence, speed can be extracted
by using various image processing techniques. Though camera based systems are more accurate
than loop based systems and do not require lane discipline they have several disadvantages. Their
performance is unsatisfactory in foggy conditions of poor visibility. Other environmental condi-
tions such as light re
ected from wet pavements and shadows a ect the performance of Video image
processors. Large vehicles can obscure smaller vehicles. Camera based systems are expensive to
install and maintain as they require quite a lot of dedicated hardware and software. Chances of
theft are also higher for such systems.
[attachment=26716]
Introduction
The road network of any city is its lifeline and so monitoring this vital infrastructural resource
is important. The problem of trac jams and congestion are faced by most major cities of the
world. For managing trac, a city's administration should have both realtime and historical data
about the trac conditions prevailing on the road network. This information can be used for
quick reaction measures, such as, changing the timings of trac lights and advising commuters
to take alternative routes through public broadcasts. In the long run, however, this information
can be used to plan a better road network by identifying areas of frequent congestion and building
alternative routes.
Apart from managing trac on the roads, maintaining the road infrastructure in good condi-
tion is necessary. Municipalities generally have tight budgets. In developing countries like India,
funds are even scarce. Hence, what the authorities want to know is where, and to what extent is a
road damaged. This would enable them to take preventive measures before further damage occurs
or prioritize repair work based on the severity of damage. It is worth noting that damaged roads
with several potholes also lead to choking of trac and cause accidents.
Problem Statement
Our goal is to build a trac and road monitoring system for intelligent route planning, road usage
and maintenance that ful lls the constraints imposed by the Indian scenario. This system should
work under varied road conditions, chaotic, dense and unstructured trac and a large variety of
vehicles. It should be cost e ect, easy to deploy (no need to dig or build overhead structures) and
require minimal maintenance. We should avoid the need for specialized equipment. In order to
meet these somewhat con
icting requirements we are willing to be content with system that does
an approximate, aggregate trac analysis and near realtime reporting. We do not want a explicit
count or classi cation of vehicles but rather some information through which we can deduce the
state of trac on a road segment. Hence, we are willing to trade-o accuracy of reporting with ease
of deployment. We want to build a road monitoring system that is able to better quantify a road
anomaly. Thus, our e orts will be to try nd out ways to report severity, intensity or dimensions
of a pothole or a damaged road segment.
Magnetic Loops
Magnetic loop is a technology that has been used for vehicle detection and trac control for the
past few decades. These devices are installed inside each trac lane and act as counters, counting
vehicle passing over them. Some variants of the magnetic loop have been used to classify vehicles
as well. The loop is a continuous run of wire which is buried inside a trac lane. The ends of
the loop wire are connected through a loop extension cable to the vehicle detector. The detector
powers the loop causing a magnetic eld in the loop area. The magnetic
ux linked with the loop
changes whenever a metal object, such as a vehicle, moves over the loop. The detection scheme
of loops is based on this principle. The change in
ux is sensed by the detector which forces a
normally open relay to close. The relay remain closed until the vehicle leaves the loop.
Camera Based Systems
Camera based systems are able to detect, count and classify vehicles. These systems use video im-
age processors to identify vehicles and their trac
ow parameters by analyzing imagery supplied
by video cameras. Images supplied by cameras are digitized and then series of image processing al-
gorithms are applied on them. Information about vehicle passage, presence, speed can be extracted
by using various image processing techniques. Though camera based systems are more accurate
than loop based systems and do not require lane discipline they have several disadvantages. Their
performance is unsatisfactory in foggy conditions of poor visibility. Other environmental condi-
tions such as light re
ected from wet pavements and shadows a ect the performance of Video image
processors. Large vehicles can obscure smaller vehicles. Camera based systems are expensive to
install and maintain as they require quite a lot of dedicated hardware and software. Chances of
theft are also higher for such systems.