Author(s): Hofma, A.F. (2023)
Abstract:
Bicycling is a major form of transportation for people in The Netherlands. However, traffic incidents and collisions, including single crashes, have increased. Modern cars are equipped with technology, such as lane assist, to make travel safer. However, bicycles have not caught up. In this paper, we will look into implementing similar lane assist technology, adopted for bicycles. The technology is based on Computer Vision and Image Segmentation. Using self-recorded data sets, the proposed solution achieved a 95% accuracy for detecting road surfaces, and a 68% accuracy for detecting bicycle paths. Using this, the location of the bicycle can be tracked. As a result, the proposed technology can be used as a basis for developing a bicycle lane assist.
Document(s):
Hofma_BA_EEMCS.pdf