Tire Track Identification: A Method for Drivable Region Detection in Conditions of Snow-Occluded Lane Lines
2022-01-0078
03/29/2022
- Event
- Content
- Today’s advanced driver-assistance systems (ADAS) commonly utilize the camera sensor to provide the benefit of increased safety to the driver. This is done by using computer vision to detect two key environmental features: lane lines and objects (e.g., vehicles, pedestrians, bicycles). Lane detections are vital for safety features such as lane departure warnings, lane-keeping assist systems, and lane-centering systems. However, when subjected to adverse weather conditions, either heavy perception of rain or snow or occlusion of lane lines due to rain or snow, the lane detection algorithms are no longer capable of detecting the lane lines. Hence, the ADAS feature is no longer providing the benefit of increased safety to the driver. The performance of one of the leading computer vision system providers was tested in conditions of variable snow coverage on the road, causing occluded lane lines, using data collected in the 2020-2021 winter in Kalamazoo, Michigan. The results show that this computer vision system was only able to provide high confidence detections in less than 1% of all frames recorded. This is an alarming result, as 21% of all crashes in the U.S. are weather-related. To increase the capabilities of ADAS in conditions of snow-occluded lane lines, a tire track identification system was developed. The dataset used was collected using the Energy Efficient and Autonomous Vehicles lab’s (EEAV) research vehicle platform from Western Michigan University. This dataset was then used to train a machine learning model to detect the tire tracks in an image. This system achieved high confidence detections of tire tracks in 83% of all frames of which tire tracks were present, an 82% increase in detections than the leading computer vision system provider.
- Citation
- Goberville, N., Kadav, P., and Asher, Z., "Tire Track Identification: A Method for Drivable Region Detection in Conditions of Snow-Occluded Lane Lines," SAE Technical Paper 2022-01-0078, 2022, .