Estimation of Kinematic Motion from Dashcam Footage
This work addresses the challenge of estimating vehicle motion from dashcams for researchers and applications in automotive safety, but it is incremental as it builds on existing methods with new data and tools.
The paper tackles the problem of predicting a vehicle's kinematic motion from dashcam footage by using ground truth data from the vehicle's on-board network and a synchronized camera over 18 hours of driving. It results in neural network models that quantify accuracy for predicting speed, yaw, and lead vehicle presence with relative distance and speed.
The goal of this paper is to explore the accuracy of dashcam footage to predict the actual kinematic motion of a car-like vehicle. Our approach uses ground truth information from the vehicle's on-board data stream, through the controller area network, and a time-synchronized dashboard camera, mounted to a consumer-grade vehicle, for 18 hours of footage and driving. The contributions of the paper include neural network models that allow us to quantify the accuracy of predicting the vehicle speed and yaw, as well as the presence of a lead vehicle, and its relative distance and speed. In addition, the paper describes how other researchers can gather their own data to perform similar experiments, using open-source tools and off-the-shelf technology.