Algorithms for Vehicle Classification

Project summary:

This project will eventually enable Mn/DOT to collect vehicle classification distributions in an automated and flexible way. We plan to use CCD cameras, video cameras, and infrared sensors in order to take sequences of images at several state-aid streets and highways. The proposed work will study the feasibility of a system that analyzes these sequences and provides useful vehicle class counts. This information can then be used in the accurate design of pavements resulting in significant gains in quality and cost.