Currently, most camera calibration methods for traffic scenes are based on vanishing points and road geometry markings with simplified camera models, which can only be applied to scenes containing straight roads. However, in practical applications, cameras are usually installed with roll angles and scenes containing curved roads, to which existing methods are not applicable. To solve the above problems, we propose a novel optimization approach for camera calibration in traffic scenes, which can be applied to curved road scenes and predict camera roll angle. Firstly, a camera space model with a camera roll angle is established for image rotation. Secondly, vehicle trajectories are extracted for the best vanishing point by a parallel coordinate system and diamond space. Vehicle trajectories are also used to obtain calibration regions for extracting road markings and edges. The road markings, edges, and the best vanishing point obtained by the above two steps automatically are more accurate and stable, especially for curved road scenes. Based on the road markings and the best vanishing point, initial calibration can be conducted. Finally, by extracting redundant markings in the calibration region, the non-linear constraint of redundant markings on the road is proposed to obtain optimized calibration parameters and predict the camera roll angle. Through experimental validation on the public dataset BrnoCompSpeed and highway scenes, the proposed approach can achieve better calibration results in both straight and curved road scenes with the mean calibration error reduced by 30% compared with the previous calibration methods.

Monocular 3D vehicle localization is an important task for vehicle behaviour analysis, traffic flow parameter estimation and autonomous driving in Intelligent Transportation System (ITS) and Cooperative Vehicle Infrastructure System (CVIS), which is usually achieved by monocular 3D vehicle detection. However, monocular cameras cannot obtain depth information directly due to the inherent imaging mechanism, resulting in more challenging monocular 3D tasks. Currently, most of the monocular 3D vehicle detection methods still rely on 2D detectors and additional geometric constraint modules to recover 3D vehicle information, which reduces the efficiency. At the same time, most of the research is based on datasets of onboard scenes, instead of roadside perspective, which is limited in large-scale 3D perception. Therefore, we focus on 3D vehicle detection without 2D detectors in roadside scenes. We propose a 3D vehicle localization network CenterLoc3D for roadside monocular cameras, which directly predicts centroid and eight vertexes in image space, and the dimension of 3D bounding boxes without 2D detectors. To improve the precision of 3D vehicle localization, we propose a multi-scale weighted-fusion module and a loss with spatial constraints embedded in CenterLoc3D. Firstly, the transformation matrix between 2D image space and 3D world space is solved by camera calibration. Secondly, vehicle type, centroid, eight vertexes, and the dimension of 3D vehicle bounding boxes are obtained by CenterLoc3D. Finally, centroid in 3D world space can be obtained by camera calibration and CenterLoc3D for 3D vehicle localization. To the best of our knowledge, this is the first application of 3D vehicle localization for roadside monocular cameras. Hence, we also propose a benchmark for this application including a dataset (SVLD-3D), an annotation tool (LabelImg-3D), and evaluation metrics. Through experimental validation, the proposed method achieves high accuracy with $$A{P_{3D}}$$ A P 3 D of 51.30%, average 3D localization precision of 98%, average 3D dimension precision of 85% and real-time performance with FPS of 41.18.