Intelligent Traffic Rule Violation Detection System
Smart traffic Rules Violation Detection System That minimizes The chances of Negligence Of The Road safety Rules
Overview
If we browse through the news, we will find the incidents of road accidents every other day. If we talk about the specific NewYork city, in the year 2020, the officially registered road accident cases count was 456. Among them, 217 were fatal accidents. The count crossed 561 road accidents and 272 fatal accidents in the year 2021. The increased count drew the attention of NewYork city police and their research showed that the primary reason for the road accidents was overspeeding and negligence of the road safety rules. So, the NewYork city police came up with an idea to bring a solution that can detect the speed, and number plate of the vehicle and detect the violated rule to generate the e-memo with a penalty amount that the vehicle owner should pay.
Client Requirement
Newyork City Police reached out to Alibaba Systems with a unique requirement of developing an AI-integrated smart traffic monitoring system that can detect traffic rule violations accurately and generate the e-challan after identifying the vehicle number plate. The client wanted to integrate the existing city CCTV cameras with the smart system to minimize the efforts of reinstalling the CCTV cameras all over the city. The client wanted a system that can accurately detect the traffic rules violating vehicles, whether it could be a car, motorbike, bus or even a truck. The system should also detect the vehicle number plate and generate the e-memos for the detected vehicle.
Solutions
- The smart traffic rule violation detection system development was a challenging task. So, our development team came up with an AI integrated solution that can monitor the vehicle traffic and capture the vehicle number plate through the OCR technology in real-time. The solution uses the existing installed camera system and enhances the results the client used to get from the previous rule violation detection system with an outstanding accuracy rate.
Challenges
- Vehicle speed detection was a crucial parameter for the system as the city CCTV cameras are installed at different heights and angles. So, our team of developers did some brainstorming and came up with a customized geometry mathematical solution that can calculate the speed of the vehicle considering all the parameters of the camera.
- People all around the city wear various head accessories such as caps, turbans, face-covering and so on. So it was hard for the system to differentiate between head accessories and helmets. We made our ML more accurate by training them to distinguish between people wearing a helmet and other head accessories.
- The existing CCTV cameras are used to cover the entire range of the city road including the side road traffic that used to create glitches in the system. So, we designed a feature that allows the admin to draw the boundaries of the road map. The cameras will only monitor traffic inside those boundaries of the road map.
End-Result
- The smart system can detect the helmetless motorbike drivers and capture their vehicle number plate to generate the e-challan.
- The smart system can check the speed of the vehicle in real-time and can generate the e-memo for the overspeeding of the vehicle.
- The smart system can also capture the 4-wheeler vehicle driver's number plate if they are driving without the seatbelt.
- The smart system can identify the vehicle and categories them into various sections such as cars, motorbikes, buses, trucks and so on.
- The system categorizes the violated rules by defining different color schemes for the different rules to make it more user-friendly for the admin.
- The system provides end results with details such as vehicle number, vehicle type, speed of the vehicle and so on.
- The admin can add, update and delete the cameras from the system, manage the generated e-memo listing, and view the details of the vehicle owner.
- The admin can also view the detailed analytics based on violated rules, rule violation time frame, highest and lowest violated rules and even compare the generated data with one another that can help them with decision making
- The admin panel can also define the start and end lines of the road map to define the road map boundary for the system to detect the moving traffic of a specific area.
Technostack
Get In Touch With Us.
One World Trade Center, 285 Fulton Street suite 8500, New York, NY 10007, United States1
alibaba.softwaresolution
info@alibaba.com
