Recognition of Highway Workzones for Reliable Autonomous Driving

In order to be deployed in real-world driving environments, self-driving cars must be able to recognize and respond to exceptional road conditions, such as highway work zones, because such unusual events can alter previously known traffic rules and road geometry. In this paper, the researchers present a set of computer vision methods that recognize, through identification of work zone signs, the bounds of a highway work zone and temporary changes in highway driving environments. Through testing using video data about highway work zones recorded under various weather conditions, their approach was able to perfectly identify the boundaries of work zones and robustly detect a majority of driving condition changes. In addition to these tests, they evaluated, using a mock work zone setup, the usefulness of their work zone recognition systems’ outputs for safe-guarding a self-driving car.