Pothole Detection Using Deep Learning Methods

This work aims detect potholes present on roads. It a pothole detection system alerts drives to impending road irregularities enabling proactive avoidance manoeuvres both driver and concerned authorities using buzzer sound and mail notification respectively. By alerting the driver in advance, the risk of accident collisions can be significantly mitigated, It makes use of advanced YOLO Algorithm which is used for object detection. It also gives a real time location of the pothole using Google Maps. The existing systems are capable of detecting the potholes but do not give real time location to the concerned authorities, thus the risk of accidents still remain. The proposed system helps to detect the presence of pot holes, alert the driver and the concerned authorities to repair them and also provide accurate location of the pot hole. This technology can be further expanded to develop vehicles equipped with advance road irregularity detection system.