Deep Learning Based Vehicle/Pedestrian Detection & Tracking System
Image Vision, Deep Learning
Real time detection and tracking of moving objects has the capability to revolutionized the efficiency of transportation system. Adaptive Traffic Signal Control, Real Time Ramp Metering, Traffic Signal Synchronization, Cooperative Control, Speed Advisory all can be derived effectively if, we can gather precise and accurate traffic data (detection) along with their microscopic characteristics (tracking). Traffic Genius Application has the capability to gather this information from conventional source (video camera) by using deep learning algorithm.
We have developed multiple object detection/segmentation algorithms which allowed for the detection/localization of pedestrians, vehicles, street signs, street signals, and various objects which may be present on roads and in the vehicle. Tracking in the temporal and spatial domain allows us to monitor the objects activity and allow us to analyse and forecast the behaviour of said entity. This reliable source of information is able to assist in modern world controller design.