The Mechanism of Surveillance System Detection of Bike Riders without Helmet & Triple Riding By Using YOLO Algorithm in Machine Learning
DOI:
https://doi.org/10.17762/msea.v72i1.2385Abstract
Helmets play a crucial role in ensuring the safety of motorcycle riders. They provide protection in the event of accidents and are required to be worn in many countries. However, despite the regulations, some riders choose not to wear helmets or wear them incorrectly. In recent years, there have been numerous studies focusing on traffic analysis, including vehicle detection, categorization, and helmet detection.Computer vision technologies have been utilized to develop intelligent traffic systems. These technologies employ techniques such as background and foreground image detection to distinguish moving elements within a scene and extract relevant features. To categorize the detected items, computational intelligence methods, particularly machine learning algorithms, are commonly employed.Machine learning, a subfield of artificial intelligence, involves training models that can learn from provided inputs during the training phase. In the context of object detection applications, machine learning techniques are used to create mathematical models using sample data known as "training data," enabling the generation of predictions or decisions.By training a helmet identification model with a specific dataset, it becomes feasible to accurately identify riders without helmets. Additionally, the license plate of the rider can be cropped and stored as an image based on the recognized classifications. An optical character recognition (OCR) model can then be applied to the image to recognize the text and output the plate number in machine-readable text format. This entire process can be implemented in real-time using a webcam.The real-time implementation of such a system relies on utilizing a dataset as input for training the models. With the appropriate dataset and trained models, this application can effectively detect helmet usage and read license plate numbers in real-time scenarios.