Smart Traffic Signal

Project Title

Smart Traffic Signal

Project Area of Specialization Artificial IntelligenceProject Summary

Traffic monitoring and controlling has always been a challenge. The exponentially increasing vehicular traffic has

led to many issues ranging from traffic congestion to increased road accidents. Improved traffic density estimation would help to curb the traffic before it becomes critical problem. Vehicular traffic has various issues that make it difficult to be monitored. It might be the climatic changes, fog, smog or 24 hours electricity supply for closed-circuit television (CCTVs) to function uninterruptedly to capture the video. Vehicular traffic estimation can also be a challenge due to the pedestrian traffic present on the roads. Vehicle counting is also a hectic task for real-time applications, where processing each and every frame would not be a feasible task to perform. There have been various methods inculcated to monitor and control traffic. Earlier, landmark method has been used for the same . The method is found to be

inefficient for the humongous traffic currently present on the roads. Then, magnetic loops came into picture. Magnetic loops could not completely solve the problem due to their high maintenance cost and not being able to bear heavy vehicular weights. Sensors have been widely used for a very long period of time for traffic density estimation. But sensors have their own limitations. They perform low level sensing

[10].Generally, sensors have a limited range of operations. Conventional methods had their cameras implanted on ground level. Elevating these cameras gave a wider scope of operation for the traffic detection and estimation. Recently, mobile systems are widely used to collect the vehicular count

using Global Positioning System (GPS) through satellites that collect information periodically. Still the non-linear behavior of the traffic cannot be completely analyzed by any of the methods mentioned above. Therefore, a framework has been introduced that would help in increasing the computational speed of the existing system and would also help in improving the traffic management for intelligent visual surveillance

Project Objectives

The goal of this project is basically to try to controls and maintains the traffic signals smartly.Our project will mainly focus on the following objectives:

1. By this we can help to our country by solving one of the biggest problems.
2. We try to monitor and control the traffic by continuously capture images with the help of CCTV video camera which will be fitted on Signals. By taking real time picture of traffic we can reduce the flow of traffic jam.
3. It can eliminate the extra waiting of traffic according to traffic situation.
4. Smart Traffic Signals can help to curb the traffic before it becomes critical problem.
5. Develop a system that automatically allots time to traffic signals according to the density of traffic.
6. It can provide high-quality image information efficiently and stably.
7. It is easy and economical to install video cameras. Besides, it would never damage the road, nor would it block the traffic.
8. With the fast development of computer vision and digital image processing technology, video based traffic flow detection system has become increasingly robust, Realtime and intelligent.
9. Our system give priority it’s very effective for emergency services.
10. We try to sense the behavior of traffic which to mostly appears, so system can reduce the traffic before that sad time.

Project Implementation Method

The Design Phase of the project started in early July, and we will continue working on the following aspects:

1. Design Person detection techniques

We will design Person Flow Detection system which help us to detect moving person using CCTV cameras

2. Design person detection system

We will design person’s face detection using AI and image processing algorithm. When system recognize face then it will check that person in our database.

3. Design system that will store person’s history

We will design the database that will store the time and place of the person which will help in security and education purpose to eliminate the time-wasting factor of student.

4. Functional Features

• Object Recognition from Camera (Video)
• Detection of Movement.
• Thresholding of Object
• Track record of that person
• which route he use for destination
• which people he spends time most
• where he spends its time
• detection of unknown person

2.2 Testing

We use TDD Test Driven Deployment for develop this system, we can collect the real possible cases that will be possible and then write test cases accordingly. Once test cases will be done we start development according to it.

During the development process, unit testing will be done to ensure all modules are built correctly. System integration testing will be done after we have built all the components and combined them into the application. We will test the database, the algorithms and the user interface.

1. Testing We use

Integration Testing

Stress Test

Black Box & White Box

Unit Test

1. Test the Database

To test the database, we will check all the tables, written queries and procedures. Testing can be performed in web application and database can be used in the application like SQL.

Our Database testing basically include the following:

1. Data validity testing.
2. Data Integrity testing
3. Performance related to data base.
4. Testing of Procedure, triggers and function.

Test Camera Angle
This is important to check the actual data of vehicles on the road, first we need to set camera properly so we get the actual picture of traffic we want.

Test Custom API

API testing perform during the development. We test our API by stress testing, we put some request and check the results.

Test Images

To test the Images, we will check the objects in image by using image processing and when we get any object in the image check its accuracy.

Test Algorithm

we test every implementation of an algorithm in that way: take an input, calculate by own hand expected output, and compare it to the output the algorithm provides.

Benefits of the Project

Vehicular traffic congestion is an increasingly growing problem in this modern world. The increase in vehicles purchased per year in no way reduces the number of vehicles on our roads. There is therefore the need to devise a system to ensure smooth flow of vehicles, especially at intersections. Standard traffic lights, with fixed intervals between light changes, have helped reduce this issue over the years. However, the increase in the number of vehicles on roads, especially during rush hours, has rendered the standard traffic lights incapable of efficiently and effectively reducing traffic jams. The motion of the vehicle is one of the basic parameters for identification of the flow of traffic on roads. The traffic flow on the roads can be basically categorized into heavy, medium and low traffic. Majorly thresholds that are used to correctly classify the traffic in any frame. Background subtraction, edge detection, optical flow estimation, computer vision filtration techniques, closure operation are some techniques that are combined by various researchers and used to almost classify the nature of vehicular traffic in a frame. We are presenting an autonomous vehicle density-based traffic control system.
 

• By this we can help to our country by solving one of the biggest problems.
• We try to monitor and control the traffic by continuously capture images with the help of CCTV video camera which will be fitted on Signals. By taking real time picture of traffic we can reduce the flow of traffic jam.
• It can eliminate the extra waiting of traffic according to traffic situation.
• Smart Traffic Signals can help to curb the traffic before it becomes critical problem.
• Develop a system that automatically allots time to traffic signals according to the density of traffic.
• It can provide high-quality image information efficiently and stably.
• It is easy and economical to install video cameras. Besides, it would never damage the road, nor would it block the traffic.
• With the fast development of computer vision and digital image processing technology, video based traffic flow detection system has become increasingly robust, Realtime and intelligent.
• Our system give priority it’s very effective for emergency services.
• We try to sense the behavior of traffic which to mostly appears, so system can reduce the traffic before that sad time.

Technical Details of Final Deliverable

We are creating a demonstration of this all process, we will detect traffic congestion from pictures of the roads from all intersections by the help of cameras built on the traffic pole and then allocate the time for each intersection separately. In this demonstration, we will create a road with the help of acrylic sheets, the black rexine and traffic signal. We will purchase toy cars for the presentation. The product will be able to detect the congestion from all four intersections and then allocate time according to it. This product will work in a round-robin manner until the emergency service appears on any of intersection. After the appearance of emergency service, the algorithm of time allocation ie. a round-robin will be substituted by priority based algorithm. All of the above workings will be controlled by a single or two raspberry-pis. We will use the Raspberry Pi 3 B +. The Raspberry Pi will work on code which will be written in python. Weights those are trained by us for emergencies vehicles and normal vehicles and python code file for Image detection will be separated for emergency vehicles detection and cars. The code file for the time allotment algorithm and traffic lights controlling code will also be in separate files.

Final Deliverable of the Project HW/SW integrated systemType of Industry IT , Transportation Technologies Artificial Intelligence(AI), Internet of Things (IoT)Sustainable Development Goals Sustainable Cities and CommunitiesRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	79976
Original Raspberry Pi 3 Model B Plus	Equipment	4	6499	25996
Logitech C922 Pro Stream	Equipment	4	11000	44000
128 GB EVO Plus Micro SD Card for Pi memory	Miscellaneous	4	1200	4800
Pi adapter & Case	Miscellaneous	4	1000	4000
paints	Miscellaneous	1	180	180
Road cover	Miscellaneous	4	250	1000