Answers are grounded in this paper and cite their sources.
There are many potholes present on the road.This can lead to major accidents.Every year around 3597 people die due to these potholes.To tackle and detect such potholes we have come with this project.The goal of this project was to create a pothole detection system that is specifically designed to detect potholes.This model was developed using the YOLO (You Only Look Once) algorithm for real time object detection.It is a pretrained model which detects the pothole using YOLO v4.Previously sequential CNN (Convolution Neural Network) Algorithm was used but later we figured out after a comparative analysis that YOLO gave better results in real time.A GUI (Graphical User Interface) was added to the model so that we can simulate the model using the start and stop button.This system consists of a camera that, when activated, extracts images from live camera recording in order to detect potholes.It will display potholes in real time, and the pothole will be highlighted with boxes, as seen in real-time object detection systems.This system provides us with an accuracy of 80-85%.The system can be integrated with traffic management systems, providing real-time alerts to drivers about the presence of potholes on their routes, allowing for timely evasive actions.
No takes yet. Share an insight, caveat, or question.
A 2024 study studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: