Deep Learning Model Regression Based Object Detection for Adaptive Driving Beam Headlights

As the world moves toward automated driving (AD) continues, the future of adaptive driving beam headlights (ADB), is quickly coming into focus. Engineers, developers and designers are researching hard to identify the most effective combination of components to meet driver requirements for safety and visibility. ADB is a technology used in automotive headlight systems that automatically adjusts the beam pattern of the headlights to provide the best visibility for the driver while also reducing glare for oncoming drivers. The system uses cameras, sensors, and algorithms to detect the presence of other vehicles on the road and adjust the headlight beams accordingly. This allows the driver to have the highest level of visibility while minimizing the risk of dazzling other drivers. ADB is available on many vehicles including Europe, Asia & Middle East. Adaptive capabilities help reveal critical objects such as lane markings, pedestrians, and oncoming cars while avoiding using full high beams that might temporarily blind an oncoming vehicle driver. However, designing and developing a solution for the real road conditions is time-intensive, expensive, and complex. Hence, there is a requirement for adaptive driving beam headlights to detect the oncoming vehicles to reduce the glare for oncoming vehicle drivers. The detection solution needs to be fast, accurate and easy to integrate with automotive vehicular systems. This paper reviews various detection techniques that can be used in implementing adaptive headlamps and application of the Machine Learning technique to predict fast and accurate object detection.