Motorbike Safety Rule Violation Detection System

A statement issued by Rescue 1122 revealed that nearly 1.3 million people die annually and over 50 million suffer injuries leading to disabilities due to road traffic accidents. The most common type of transportation in Pakistan is a Motorbike. Yet unfortunately, Pakistan is one of the countr

Project Title

Motorbike Safety Rule Violation Detection System

Project Area of Specialization

Artificial Intelligence

Project Summary

A statement issued by Rescue 1122 revealed that nearly 1.3 million people die annually and over 50 million suffer injuries leading to disabilities due to road traffic accidents.

The most common type of transportation in Pakistan is a Motorbike. Yet unfortunately, Pakistan is one of the countries where Motorbike Safety Rule violations of Helmet and exceeding the number of permitted riders have always existed at their peak even when the government is trying its best to make it minimum. Over 60% of deaths in Bike crashes were for people who were not wearing helmets and exceeding riders.

PMVO 1969, 89-A, Pakistan motorcycle helmet law mandates that: “Rider to wear a helmet. - No person shall drive, or ride the pillion seat of, a two-wheeled motor vehicle except when he is wearing a crash helmet”.

In the era of modernization, (to reduce human labor and add features that focuses on the life-threatening violations of bikers in the existing safe city system) Our project aims to ensure the safety of bikers by detecting the violation of helmets and exceeding riders' limit on a bike.

Using Image processing, this work will identify the objects that reside in the images taken out from the video surveillance as multiple frames. Deep learning techniques will be used for image or pattern identification along with Visual Geometry Group (VGG), which is mainly used for object detection. If a bike rider is traveling without a helmet and exceeding the rider's limit (more than two persons), the image of the number plate of the Bike will be captured. The system uses a pure machine learning algorithm for image processing. Identification of the motorcycle can be made in five steps: image capturing, pre-processing of an image by finding the errors, image recognition, and feature extraction.

Project Objectives

• Making bikers Travel safely by enforcing them to follow traffic rules.
• To detect and recognize helmet of rider’s on bike.
• To save maximum lives by enforcing rider's limits on a bike.
• To provide proper and efficient Online traffic rules violation detection and fine system.

Project Implementation Method

The system proposes a system for helmet detection on motorcyclists for safety and surveillance using the deep learning YOLOv4network. OpenCV was used to process the video as well as detect the motorcyclists based on aspect ratio and reference lines. The tests on a large set of helmet positive images and tripling showed the system to detect about 80% of all vehicles correctly. The system can befurther integrated with a license plate detection system to allow law offenders to be tracked and then penalised. Furtherimprovements can be made to the accuracy of the system by using a more extensive database

Benefits of the Project

Motorbike Safety Rule Voilation Detection System will recognize a motorcycle and determine whether or not the rider is wearing a helmet and either it is exceeding a rider-limit or not. These are following features listed below that system will provide:

• A system which will detect bike riders with and without helmet with reasonable accuracy.
• A system  which will detect number of bike riders (exceeding riders limit) with reasonable accuracy.
• A system will detect the number plate of a bike.
• A system will generate a report that will show the numbers of bike-rider without helmet and riders exceeding riders limit on a bike.
• A system will  be able to create a challan for riders violating rules.
• An interface for an admin to manage a challan and related information.

Technical Details of Final Deliverable

Using Image processing, this work will identify the objects that reside in the images taken out from the video surveillance as multiple frames. Deep learning techniques will be used for image or pattern identification along with Visual Geometry Group (VGG), which is mainly used for object detection. If a bike rider is traveling without a helmet and exceeding the rider's limit (more than two persons), the image of the number plate of the Bike will be captured. The system uses a pure machine learning algorithm for image processing. Identification of the motorcycle can be made in five steps: image capturing, pre-processing of an image by finding the errors, image recognition, and feature extraction.

• Image Processing
• Label Img
• YOLO Model
• Vs code
• Colab
• Anaconda
• Open-Cv Python
• Camera
• Mysql
• Amazon Services

Final Deliverable of the Project

Hardware System

Core Industry

IT

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Good Health and Well-Being for People, Industry, Innovation and Infrastructure

Required Resources

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