Smart Traffic Rule Violation Detection System
AI Integrated Smart Traffic Rules Violation Detection System to Enhance the Probability of Human Life.
Traffic rules violations have been so normal that we don't even consider them a problem. Most of us just pay a bit amount to the police officer whenever we get caught in action, and it's not something big. But not everyone is turning a blind eye toward these concerns. One of the major cities of India has implemented a smart traffic violation detection system that can detect the speed of the Two-wheeler and four-wheelers and sends them an overspeeding e-memo. Along with that, the system also detects the 2-wheeler rider driving without a helmet and the 4-wheeler driver driving without a seatbelt accurately. All the drivers of Ahmedabad city who violate speeding or safety rules while driving will get the e-memo automatically without any human interference. It will minimize the chances of road accidents and enhance the riders' safety.
Client Requirements
The Ahmedabad City Police wanted a system that can detect the speed of the vehicle, type of the vehicle, and number plate in real-time to determine whether the vehicle is overspeeding or not. Along with that, the system should also detect the 2-wheeler drivers driving without helmets and 4 wheeler drivers without a seatbelt accurately through the usage of AI and ML technologies.
As there are various cameras installed all over the city, the client wanted to use the existing system and cameras for smart traffic rule violation detection. So, Hyperlink InfoSystem came up with a customized AI embedded solution that can fulfill every requirement of Ahmedabad Traffic Police and minimize human efforts.
Features
The Traffic Rule Violation Detection System Features.
Process Automation
The smart system automates the basic work of traffic rules violation activity such as capturing the vehicle license plate number along with the speed of the vehicle and other needed driving security measures that used to demand manual efforts in the previous system.
Existing System Usage
The smart system uses the existing cameras of the city police to automate the traffic rule violation detection activity using advanced technologies such as Artificial Intelligence and Machine Learning.
Number Plate Detection
The system uses OCR technology to identify the license plate number and extract the numbers from it. It will help the admin to get the precise number plate of the vehicle.
Vehicle Type Detection
The system can classify the vehicle type using AI & ML technologies into various categories such as car, motorbike, bus, truck and so on.
Vehicle Speed Detection
The system uses the same existing installed IP camera to detect the speed of the vehicle. The system can detect the moment of the vehicle in the video stream to get an accurate speed estimation in KMH format using OpenCV and a deep learning algorithm.
Helmetless Detection
The System can detect the 2 wheeler riders who are not wearing helmets through the object detection model. The system can detect the license plate number of the helmetless rider in real-time.
No Seatbelt Detection
The system can detect the category of the vehicle. The system can detect the license plate number of the 4-wheeler after identifying whether the driver is wearing the seatbelt or not.
Accurate Representation Of The Data
The system can represent the collected data accurately through various color schemes such as green, red, orange, and so on that can help the admin to identify the violated rule along with the vehicle details such as vehicle number, vehicle type, and speed of the vehicle.
Admin Panel
The Admin panel has a lot of handy functionalities. As per the client requirement, we have developed a dashboard for the admin where he can manage the users, posts, categories, content violations and restrictions along with detailed app analytics about application usage to user engagement.
Dashboard
Admin can view the detailed analytics for various traffic rules such as overspeeding, helmetless vehicles and various others based on month and year. Admin can even compare the data with one another to get accurate details.
Manage City Camera
Admin can add, delete, activate and deactivate multiple city cameras to the smart system.
Manage Listing
Admin can view the listing of the traffic rule violating vehicle along with the various details such as vehicle type, vehicle number, vehicle owner name & address, license number, violated rule, date and time of the rule violation, previous unpaid challan amount, current rule violation amount and so on.
Memo Verification
Admin can manually approve and reject the generated e-memo to enhance the accuracy of the working system.
Admin Profile Creation
Admin can create various profiles based on the users' roles that can structure the flow providing limited system access based on their role. It can enhance the safety and security of the system from the admin side. Even with admin access, the admin can not access the entire system if they are not suitable candidates to access specific system functionality.
Boundary Creation
Admin can draw the start and end line of the boundary to create the road boundary. The system will detect the traffic violation rule only within the road boundary.
Project Approach & Results
Well, we all know that we have never been serious about the traffic rules whether it could be driving without a seatbelt or using mobile phones while driving. That has caused serious road accidents throughout these years. To take a step forward toward road safety, Ahmedabad City Police wanted a smart system that can detect the number plate of traffic rules violating vehicles.
Our core aim for the development of this system was to make the traffic system more accurate using advanced technologies that can minimize human efforts. To begin with, we analyzed the concept hypothetically and created a file with the complete details covering all these points, diagrams, scenarios, problems, flow charts, and SRS for the entire workflow and planned a smart traffic rule violation detection system development process. After finalizing the document, In the third phase, based on the Ahmedabad City Police's requirements and our findings, we defined smart system architectures for the system. After that, Our team of designers started working to prepare the wire-frames and design of the system. After getting approval on the final app design, our team of developers move ahead with the development phase. We created the prototype for the smart traffic rule violation detection system with all the native data. After the client's review and confirmation of the prototype our team of developers started their development process by choosing the best-fit technology for this AI integrated system.
We easily get access to previous video footage of the city traffic system as the city police have already installed thousands of cameras on almost every crossroad of Ahmedabad, and we have to use those cameras only. We started our development process using those pre-recorded IP camera footage to train the system for different camera angles, light conditions, different traffic crowd conditions, night views and so on. After that, we classified the vehicle using AI and ML technologies to implement real-time CCTV cameras. We trained the system using video recording data for accurate results.
Even though the previous system was used to detect the vehicle number plate, helmetless riders, and traffic signal rule violations, the result was not as accurate as it should be and the admin had to generate the memo for the valid vehicle owner manually. Our smart traffic rule violation detection system is more accurate and equipped with advanced technologies such as artificial intelligence and machine learning that automate the process starting from vehicle type detection to generating the e-memos of the vehicle owner. The dedication and teamwork of our team of developers led us to deliver the solution successfully, as well as within the prescribed timeline & budget. After the development phase, our QA team tested the working of the entire system before finally delivering it to the Ahmedabad City Police.
Challenges
1. Speed Calculation
The primary concern of the smart system was calculating speed. As the CCTV cameras are located at different heights and angles throughout the city. So we did some extensive research and came up with a geometry mathematical solution by which we calculated speed considering all parameters of the various cameras.
2. Helmet Detection
One of the primary requirements of the system was helmetless detection. As various people around the city prefer to wear caps, turbans, and various other face-coverings. To overcome this challenge, our team of developers trained the system with the data of real people wearing a helmet, cap face-coverings and so on for our ML to distinguish classification.
3. Boundary Detection
As the CCTV cameras cover the entire range of the city road it used to cover the side road traffic as well. So, our team of developers came up with a customized solution that can help the admin to draw the start and end lines of the city road map that can create the boundary of a portion of the area to observe the traffic of the target road.
Technology Stacks
We have used the latest technologies that can justify client requirements at the best to deliver bug-free solutions.
