RTPD-YOLO: Reconciling YoLo-V8 Model for Real-Time Potholes Detection

Indian road privilege systems employ low-quality building materials, making pothole detection and avoidance difficult. Preventing accidents requires quickly identifying and fixing potholes. Roadside potholes may damage cars and endanger drivers. This article introduces a new method for real-time pothole identification called RTPD-YOLOv8. It can identify several forms of potholes, including cracks, oil stains, patches, and stones. To begin processing, pothole videos are first transformed into picture frames. The Ultratics are applied to these frames before processing to lessen distortions. The next step is to use the pre-processed photos to detect the pothole and identify its region using the Sobel edge detector and YOLOv8. Thereafter. The proposed RTPD-YOLO will detect the potholes efficiently. The Simulation results show that Accuracy, Precision, Recall, and Fl-score were used to measure the performance of the RTPD-YOLO approaches. A total Accuracy of 99.72% is attained by the RTPD-YOLO. Also, the RTPD-YOLO model gets a 98.86% precision, 94.64 percent recall, and 91.35% Fl-score. When compared to other models, like Faster RCNN, SSD, and mask R CNN, the RTPD-YOLO network achieves a higher Accuracy range.