Lane Change For System-Driven Vehicles Using Dynamic Information

Over the years, the advanced driver assistance system (ADAS) has evolved into a mature system-controlled driving. This includes automatic parking, obstacle detection, etc. One of the important aspects that still need to successfully investigate is the automatic lane change across different traffic scenarios and the different nature of roads. While it is still easier to identify the road, the rapid change in the traffic makes it difficult to safely carry out automatic lane change in real-time. This work proposes a mechanism by first modeling the behavior of vehicles in a dynamic environment by generating fusion data from various externally mounted sensors, RADAR and cameras and then develop a visual perception of the same using multi-object tracker and tracking Kalman filter.