Efficient automatic braking in a car is the key to the success of the autonomous vehicles. For implementation of the efficient automatic braking in the vehicles, it is important to detect the obstacles in the fontal area. The obstacles can be succeeding vehicle, human being, animal or etc. Through this research paper, efficient xception model is disclosed for identification of the braking light of the succeeding car. For implementation of this model, in the first phase car is detected using Haar-Cascade modelling. In the second phase of the development, brake light of the succeeding car is detected. Signaling of the succeeding car is categorized into parking indication, left or right turn indication, off state and braking light. For the input stream, accurate indication with selection mask becomes visible with the classification. The model precisely predicts the brake light with 99% accuracy.
Exception Model for Brake Light Detection for Autonomous Vehicles
2022-06-24
1382392 byte
Conference paper
Electronic Resource
English