This technology is an automated lane marking system that identifies lane conditions using cameras mounted at the front and near the ground of an unmanned vehicle, determines the vehicle's position, moves the vehicle to the marking location based on analysis, and applies paint via internal nozzles.
Lane markings are essential for guiding drivers, requiring periodic repainting. Traditionally, this relied on manual steering or operators driving vehicles to align positions. Because external workers had to monitor the process to ensure accuracy, they were exposed to safety risks, and the process was labor-intensive.
This technology uses a front-facing camera for site monitoring, while side-mounted cameras capture road surfaces and wear levels. After filtering for color and brightness, it extracts lanes using iterative Hough transforms and RANSAC, correcting steering based on the vanishing point of parallel lanes. New markings are plotted using road blueprints and 3D maps. Surface defects are identified via contrast analysis, and line defects via scratch detection algorithms. Cracks are cleaned with a built-in vacuum before molten paint is injected. Applicable to highways, new roads, airport runways, and parking lots, it eliminates the need for flaggers and monitors, reduces lane closure times, and enables precise repainting of only worn sections.
This invention was developed with support from the Ministry of Education's Capstone Design program.
N/A