FPGA Suitability for Implementation of Pedestrian Detection algorithms for Autonomous Driving Application

An emerging field of study called autonomous driving seeks to create self-driving cars that can guarantee the safety of both passengers and other road users. Although several countries have already developed basic autonomous vehicles, more research is required for more sophisticated automation. Numerous different computer vision (CV) algorithms are needed to handle the extensive elements in an autonomous vehicle. Although many computer vision algorithms have already been built as software, they still require a stable hardware platform in order to be tested and used. Due to the many advantages it offers, Field Programmable Gate Array (FPGA) has become the most popular hardware platform. This paper addresses the significance and necessity of autonomous vehicles, highlighting one of their most difficult features— pedestrian detection. The paper also reviews the numerous advantages of FPGAs specifically for the development of autonomous driving challenging features like pedestrian detection and it also documents that how different researchers have used FPGA's as their hardware platform to construct self-driving car's features with an eye towards potential future improvements.