Smart Traffic Management System

The increase in the mismanagement of traffic and consequent congestion has arisen many problems. Considering our own city Karachi, to reducing the problems, the idea of Smart Traffic Signal Management System has suggested. By creating our own data set based on images captured from within the city, t

Project Title

Smart Traffic Management System

Project Area of Specialization

Artificial Intelligence

Project Summary

The increase in the mismanagement of traffic and consequent congestion has arisen many problems. Considering our own city Karachi, to reducing the problems, the idea of Smart Traffic Signal Management System has suggested. By creating our own data set based on images captured from within the city, the proposed project aims to make decisions for the traffic signal timings based on the vehicle densities. The current traffic controlling system is either manual or rigid. These are not efficient methods for random numbers of vehicles on roads. The purpose of this system is to regulate the signal timings based on the count of automobiles. The project would deploy on a four-way traffic signal. For surveillance, cameras will get deployed. The camera aims to send in the real-time video of the road into the raspberry-pi. It will proceed for image processing to separate image frames. Mainly, image framing and vehicle detection will assist in counting them. Machine-learning algorithms will perform the task signal controlling and vehicle detection. It will play a vital role in concluding the time for lights for each side of the four-way. Eventually, this will direct the system towards converting the fixed management one into an efficient one based on traffic density on the road.

Project Objectives

The objective of the proposed system is to improve the efficiency of the existing traffic management system of our country Pakistan. Initiating from our own city, Karachi. Through image processing and machine learning, we will be training this to work on our regional traffic system. It will assist in calculating the change of time of LEDs concerning the current traffic load. The camera will take images from the video at various time to sense the traffic by processing it. The frame of the lane will be processed using image processing techniques to evaluate traffic. The predicted traffic load on a particular road would help calculate the required time duration signal lights. The system will be intelligent and will calculate time to time and operate in a loop. The decision making will assist in preventing over waiting of vehicle in the queue of other lanes.
The main aim of the Smart Traffic Management System is to reduce the waiting time and accidents. The system consists of several parts according to its functionalities. First part is the camera, the essential part of the system. It will be used for monitoring and taking real-time video. Second is the controller, the brain of the system. It will generate a command to the traffic lights based on the calculation. The calculation would consist of comparisons of the number of vehicles on the side by side lanes, the one with maximum will be shown green signal for the decided time.

Project Implementation Method

The proposed smart traffic management system uses video data collected of the traffic of the roads of Karachi, Pakistan, from the camera, performing machine learning algorithm over the recent frame obtained from the video to estimate the number of vehicles present in a scene. Cameras installed on the opposite of the lane, beside the traffic light, will take its real-time video. At the backend, raspberry pi would be connected, which would be responsible for video processing. Raspberry pi would receive video as input from the camera of each road. Image framing would capture frames from the video at several fixed intervals.
The latest frame would be sent into a testing module created with a machine-learning algorithm. This module will detect the vehicles and number of vehicle present. Smart vehicle counting system counts less than the actual number of vehicles due to congestion and less distance between cars. But the accuracy is high due to lots of trained data. 
Based on vehicle densities, an algorithm designed with machine-learning will do the decision making. To limit the time for on and off intelligently concerning a load of vehicle on the respective lane.

Benefits of the Project

• Reduction in Traffic Jams,
• Reduction in time wastage,
• Video log of traffic would be created at the backend,
  • This will work as surveillance. Images of every second will be saved that will help in looking for anything important later on.
• Reduction in Pollution.

Technical Details of Final Deliverable

The proposed system is set to be trained in the regional traffic system. Videos of traffic from the road of Karachi would be recorded. These videos would be then used in creating a dataset of our own. The dataset would hold all the local vehicles that would make it efficient to work properly in our city.

The self-created dataset will be trained in our system which will help us detect vehicles in the video. Detection would be done using YOLO, a CNN algorithm. On detecting the automobiles, they will be counted.
To control the traffic signal, the density of the road will be used. Using a machine-learning algorithm, time for each light would be set according to the count of vehicles on the lane.
The whole process would be implemented on the Raspberry Pi. The video from the camera would be sent as an input to the Pi which will do the detection, counting and controlling process. This will eventually lead to the output that will be shown in terms of delays in lights of the signal.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Transportation

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Sustainable Cities and Communities

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com