Machine Learning and Internet of Things-based Driver Safety and Support System

Accident patterns can be determined through driving analysis. In environmental, safety, and fleet management applications, accident detection can decrease potential risks to vehicles and drivers, enhance service quality, and save money. Earlier automatic accident detection methods relied on traffic data or external sensors. However, traffic data is difficult to obtain, and external sensors might be untrustworthy and inconvenient. Since traditional techniques have been constrained by a shortage of accident data, cutting-edge technologies such as machine learning (ML) have been underutilized. The most challenging aspect of the task is detecting roadside barriers. The proposed idea is to create something that can identify barriers while driving a car. A Raspberry Pi Camera module aids in the object detection and identification process. The most effective technique for detecting image boundaries is determined by testing on an example image. This study recommends an Advanced Classifier to determine different categories. The proposed technology can identify cars, animals, people, etc. If an object is in the user's route, the system will warn them to slow down. In particular, the proposed Auto-Alert uses both CNN and LSTM models. Importantly, this requires the design and proof of concept of a credible traffic monitoring and density estimation system. The proposed work shows how important CNN and LSTM networks are to help drivers in traffic. Results show that the suggested method is more accurate and faster than the present methodology.