Traffic surveillance camera calibration by 3D model bounding box alignment for accurate vehicle speed measurement

Abstract

In this paper, we focus on fully automatic traffic surveillance camera calibration, which we use for speed measurement of passing vehicles. We improve over a recent state-ofthe-art camera calibration method for traffic surveillance based on two detected vanishing points. More importantly, we propose a novel automatic scene scale inference method. The method is based on matching bounding boxes of rendered 3D models of vehicles with detected bounding boxes in the image. The proposed method can be used from arbitrary viewpoints, since it has no constraints on camera placement. We evaluate our method on the recent comprehensive dataset for speed measurement BrnoCompSpeed. Experiments show that our automatic camera calibration method by detection of two vanishing points reduces error by 50 % (mean distance ratio error reduced from 0.18 to 0.09) compared to the previous state-of-the-art method. We also show that our scene scale inference method is more precise, outperforming both state-of-the-art automatic calibration method for speed measurement (error reduction by 86 % -7.98 km/h to 1.10 km/h) and manual calibration (error reduction by 19 % -1.35 km/h to 1.10 km/h). We also present qualitative results of the proposed automatic camera calibration method on video sequences obtained from real surveillance cameras in various places, and under different lighting conditions (night, dawn, day).

Training-free Monocular 3D Event Detection System for Traffic Surveillance

“…1) Camera Calibration: We adopt a road surveillance camera model as in [7]. According to [8], the relationship between the homogeneous coordinates of a point X =…”

“…3) Turning Events: The trigger and border conditions of turning events are |a θ | ≥ a θ,trigger and v r ≥ v turn min (6) |a θ | ≥ a θ,border and v r ≥ v turn min (7) where a θ,trigger > a θ,border . For a turning event started at t s and ended at t e , the turning angle can be calculated as…”

Traffic Danger Recognition With Surveillance Cameras Without Training Data

“…Different methods may require various forms of user inputs, such as drawing parallel lines [12], camera position [22,14], and average vehicle size [3] or speed [16]. Fully automatic calibration is also achievable according to [4,18].…”

“…The dataset provides the ground truth of distance measurement lines and speed of vehicles from Lidar sensor. It also has calibration results from various systems [19,18].…”

“…We do not use lower frame rate because the Deep-SORT model has worse performance when it is less than 25fps. We exploit the calibration results from [18] which provides vanishing points u, v acquired by manual calibration from parallel lines, along with scale factors inferred from speeds. We let the smoothing parameter δ = 0.86 according to some preliminary experiments.…”