Application of Computer Vision for Estimation of Moving Vehicle Weight

Abstract

Heavy vehicle weights need to be closely monitored for preventing fatigue-induced deterioration and critical fractures to highway infrastructure, among many other purposes, but development of a cost-effective weigh-in-motion (WIM) system remains challenging. This paper describes the creation and experimental validations of a computer vision-based non-contact WIM system. The underlining physics is that the force exerted by each tire onto the road is the product of the tire-road contact pressure and contact area. Computer vision is applied (1) to measure the tire deformation parameters so that the tire-roadway contact area can be accurately estimated; and (2) to recognize the marking texts on the tire sidewall so that the manufacturer-recommended tire inflation pressure can be found. In this research, a computer vision system is developed, which is comprised of a camera and computer vision software for measuring tire deformation parameters and recognizing the tire sidewall markings from images of individual tires of a moving vehicle. Computer vision techniques such as edge detection and optical character recognition are applied to enhance the measurement and recognition accuracy. Field experiments were conducted on fully loaded or empty concrete trucks and the truck weights estimated by this novel computer vision-based non-contact WIM system agreed well with the curb weights verified by static weighing. This research has demonstrated a novel application of the computer vision technology to solve a challenging vehicle WIM problem. Requiring no sensor installation on the roadway or the vehicle, this cost-effective non-contact computer vision system has demonstrated a great potential to be implemented.