Effect of Camera and Video Parameters on the Accuracy of Vehicle Speeds and Accelerations Calculated from Onboard Dashboard Cameras

The goal of this study was to evaluate the influence of camera and video parameters on the accuracy of speeds and decelerations calculated from onboard dashboard cameras (“dashcams”). We equipped a single test vehicle with 5 commercially available dashcams. We also equipped the test vehicle with a 5th-wheel, a brake pedal switch, and a synchronization light visible to the dashcams. We conducted 9 emergency braking tests from about 60 km/h and calculated the reference speed and deceleration from the 5th-wheel data. The brake pedal switch indicated the moment of brake application. The light synchronized the dashcam videos to the 5th-wheel and brake pedal data. The 5 dashcams had horizontal fields of view (FOV) ranging from 120 to 170 degrees, frame rates of 30 or 60 frames per second (FPS), and resolutions of 1080p, 2K, or 4K. For each camera and brake test, we extracted the video from the dashcam and calculated its linear and rotational motion using SynthEyes, a 3D motion tracking and video analysis program. Scale and location for the analyses were based on 3D measurements of the test road acquired by a laser scanner. We compared i) the SynthEyes-calculated speed during steady-state motion prior to braking to the 5th-wheel speed, and ii) the deceleration rate calculated from the SynthEyes solution during emergency braking to the deceleration calculated from the 5th-wheel data. The biases and uncertainties of the 5 dashcams and their associated FOV, FPS, and resolution were compared to provide insight into the reliability of speeds and decelerations calculated from dashcam videos.