2D空间中线扫相机的静态标定方法

Static Calibration Method for Line-Scan Cameras in 2D Space

针对当前线扫相机标定过程复杂且需外部辅助装置的问题,提出建立在2D空间中的线扫相机静态标定方法。首先,从应用角度出发,将3D空间的线扫模型推导至2D空间,建立2D空间线扫相机成像模型,设计一种新型的线扫相机标定板。接着,利用交比不变性原理对2D空间的线扫相机模型进行求解。最后,提出一种利用畸变变化量的畸变拟合策略,解决小畸变情况下畸变矫正误差过大的问题,取得了良好的畸变矫正效果。实验结果表明,所提2D空间中的线扫相机静态标定方法标定过程简便,不受现场环境限制,标定精度较高,取得0.06 mm的平均绝对误差(约1.5像素),误差标准方差稳定在0.068以下。所提方法基本满足在中低标定精度下的应用要求,具有较高的实际应用价值。

The calibration process for a line-scan camera is typically complicated,often requiring external auxiliary devices.This paper proposes a static calibration method for line-scan cameras built in 2D space to address this problem.First,from the application point of view,the line-scan camera imaging model in 3D space is adapted to derive an equivalent model in 2D space,and a new line-scan camera calibration plate is designed.Next,the line-scan camera imaging model in 2D space is solved using the principle of the invariance of the cross-ratio.Finally,an aberration correction fitting strategy is proposed,using the amount of aberration change to overcome excessive aberration correction errors,which can occur when trying to correct small aberrations.The aberration correction fitting strategy is effective,achieving a good aberration correction effect.The proposed static calibration method for line-scan cameras in 2D space is simple,is not limited by the field environment,and has a high calibration accuracy.The proposed calibration method has a mean absolute error of 0.06 mm(equivalent to~1.5 pixels)and a maximum stable error standard variance of 0.068.As a result,the proposed calibration method meets the application requirements for low and medium calibration accuracy,and is extremely promising for practical applications.