End-to-End Monocular Vanishing Point Detection Exploiting Lane Annotations
This work addresses the need for accurate online camera calibration in autonomous driving systems, though it is incremental as it builds on existing lane detection and heatmap estimation techniques.
The paper tackles the problem of costly manual camera calibration in automotive applications by introducing an end-to-end vanishing point detection method that uses lane annotations to generate geometrically consistent labels, achieving higher accuracy than methods relying on manual annotation or lane detection.
Vanishing points (VPs) play a vital role in various computer vision tasks, especially for recognizing the 3D scenes from an image. In the real-world scenario of automobile applications, it is costly to manually obtain the external camera parameters when the camera is attached to the vehicle or the attachment is accidentally perturbed. In this paper we introduce a simple but effective end-to-end vanishing point detection. By automatically calculating intersection of the extrapolated lane marker annotations, we obtain geometrically consistent VP labels and mitigate human annotation errors caused by manual VP labeling. With the calculated VP labels we train end-to-end VP Detector via heatmap estimation. The VP Detector realizes higher accuracy than the methods utilizing manual annotation or lane detection, paving the way for accurate online camera calibration.