Pedestrian detection based on TensorFlow YOLOv3 embedded in a portable system adaptable to vehicles

With the expansion of accessibility and availability of personal cars towards the population, the environment, road safety, and pedestrian safety faces greater problems. The neediest for the management of information through autonomous systems dedicated to pedestrian detection and improvement of analysis and road prevention algorithms is an important factor addressed in this paper. The purpose of this paper is to demonstrate and propose viable solutions that will help drivers to practice an efficient, safe, and event-free driving style. This paper presents a prototype under development that can avoid various traffic events, the system analyzing, and alerting the driver regarding pedestrian intentions, marking each detection separately according to the degree of danger that it constitutes for both the driver, as well as for pedestrians. This research analyzes and emphasizes that at this point everything is already focused on the paradigm from which it is possible for all these technologies to cooperate in a hybrid platform, offering a real solution to the demands of human users but also of IoT solutions.