AN AUTOMATED MACHINE-LEARNING APPROACH FOR ROAD POTHOLE AND OBJECT DETECTION IN ADVERSE CLIMATIC CONDITIONS

Abstract

Autonomous vehicles and sophisticated automotive active-safety systems mostly depend on visual information to identify and locate items like pedestrians, traffic signals, lights, and other automobiles in the vicinity. This helps the corresponding vehicles navigate their surroundings safely. However, in adverse weather circumstances, such as rainy ones, the effectiveness of object-detecting techniques may decline very noticeably. Therefore, there is a need to develop software-based solutions that can enhance object detection in varied weather conditions for long-range automotive applications. Road safety is critical in today's culture since so many lives and valuable resources are at risk. Two main sources of accidents and dangers on the road are animal interactions on or near roadways, as well as the presence of potholes and other surface flaws. To proactively address these problems and mitigate their detrimental impact on road users, creative solutions are needed. In this situation, applying machine learning (ML) and artificial intelligence (AI) techniques is a practical way to improve traffic safety. The study offers a ground-breaking approach that makes use of AI and ML technologies to solve animal encounters and potholes in roads.