Publication Date: 2021/04/16
Abstract: Navigation in cluttered and crowded environments has been an important and difficult problem in technology. This involves accurately predicting pedestrians’ movements, dynamically analysing developments in the surroundings, and adjusting the path accordingly. This paper focuses on solving the navigation problem by predicting the trajectories of pedestrians. Humans are identified and tracked using state-of-the-art object detection techniques. R-CNN and YOLO are proven to have the best accuracy and speed to perform the task. We used both social and non-social algorithms to predict trajectories of the detected pedestrians. These trajectories are used to estimate future positions of the pedestrians. Finally these positions are used to calculate the path through the environment.
Keywords: Navigation, Realtime, Trajectory Prediction, Deep Learning, Object Detection, Path planning, Long Short Term Memory Network (LSTM), Region based Convolutional Network (R-CNN)
DOI: No DOI Available
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT21APR178.pdf
REFERENCES