Towards of a modular framework for semi-autonomous driving assistance systems

Road traffic accidents are a leading cause of deaths globally and represent the main cause of death among 15–29 years olds (World Health Organization in Global status report on road safety 2015, World Health Organization, Geneva, 2015). Efforts have been made to develop and improve advanced driver assistance systems as to provide better quality of driving and reduce the number of traffic accidents. However, most of the current solutions focus on developing systems for fully autonomous cars. This study investigates a modular architecture that could potentially assist drivers in situations such as traffic jams, autonomous parking, and detection of obstacles such as pedestrians. The aim is to develop a low-cost driver assistance system that could be acquired and easily incorporated by non-technical users. It is becoming increasingly feasible to build a powerful and low-cost system because of the low cost of big-sized cameras, electronics, and processing cards. As a proof of the concept, a hardware and software solution had been developed for autonomous steering-wheel actuation that only utilizes a stereoscopic camera sensor.