相机标定的外界影响因素分析

针对相机标定结果易受外界因素干扰的问题,为了提高标定准确度,利用已有的摄像机针孔成像模型,采用自适应角点检测算法提取靶标图像中的特征点,标定结果以重投影横纵像素误差的平均值作为性能指标,对均匀光源的照度、标定图片数量以及标定靶标上棋盘格尺寸3个影响因子做了相应的分组对比实验。研究结果表明,选择亮度高的光源提升标定准确度达到38%以上;特征点数目156个时,仅需18张～22张标定图片;相对较小的棋盘格尺寸可以使得标定准确度提高50%。以上结果充分说明了光源、图片数量和棋盘格尺寸对于提高相机标定准确度具有重要意义。     

摄像机标定中的特征点提取算法研究与改进

为提高摄像机的标定精度,必须研究高精度的特征点提取方法。针对三种比较典型的标定图样:棋盘格,二维正弦条纹和高斯点阵,分析比较了相应的提取特征点算法和用于摄像机内参数标定的精度,得出基于傅里叶相位分析法的二维正弦条纹的特征点提取精度最高。并针对傅里叶方法处理二维正弦条纹不能准确提取边缘特征点的缺陷,提出了使用相移法提取正交相位以代替傅里叶方法得出正交相位的新方法。这种方法可以消除傅里叶变换滤波对点阵边缘的模糊作用,扩大了二维正弦条纹在摄像机标定中的适用范围。摄像机外参数标定实验表明该方法在处理包含边缘点的点阵时,重投影误差保持在原有水平,证明了其有效性。     

电子白板系统中参数标定方法的研究

为了避开复杂繁琐的摄像机标定过程,提出了一种基于CCD摄像机的电子白板系统参数标定新方法,实现了开机自动标定系统参数的功能。采用Harris角点检测算法成功提取了参考点的像素坐标,并直接运用射影变换交比不变性,推导出白板边框图像像素距离与实际物理距离的映射关系式,进而获得了摄像机的光心坐标。简要分析影响系统标定精度的因素,并设计出减小标定误差的方案,最后搭建实验平台标定系统参数。实验结果表明:该标定方法实现简单,应用方便,且对于单位长度为1 cm的标定块而言,标定误差小于0.04 cm。     

点阵列标定模板图像特征点提取方法

在摄像机标定过程中,提取标定模板图像特征点的精度是影响摄像机内外参数标定精度的重要因素。本文根据条纹投影三维轮廓测量实验系统的要求制作了点阵列标定模板,在图像处理边缘理论的基础上,以圆的解析特性为依据,采用坐标转换思想,引进圆系描述点、圆对应关系,运用统计学理论,提出提取点阵列标定模板图像特征点的新方法,并通过实验验证了该方法的正确性和可行性。为下一步的摄像机内外参数求解做了铺垫。     

基于傅里叶条纹分析的多摄像机标定方法

提出一种新的摄像机标定方法,该方法基于2D共面参照物摄像机标定方法和傅里叶条纹分析方法.将已知相位分布的平面二维正弦灰度调制条纹图作为平面标定靶,通过傅里叶条纹分析方法计算出两个截断正交相位分布,利用截断正交相位分布并结合二维正弦条纹图特点提取相应的图像特征点,建立像素坐标与2D平面坐标的对应关系.将该二维平面靶在摄像机成像空间中放置不同的位置,并完成相应的特征点提取,根据2D共面参照物摄像机标定方法即可完成摄像机标定.该方法利用平面相位测量的高准确度来提高标定特征点的提取准确度,从而提高标定准确度.实验对双摄像机系统进行了标定,标定后该系统对标定靶进行测量,标准偏差达到0.010 mm.     

基于ICP算法的双目标定改进方法研究

双目视觉作为一种非接触三维(3D)测量技术,其位姿标定结果的好坏将直接影响3D物体测量的精度。基于迭代最近点(ICP)算法获得两组点集之间平移和旋转参数的原理,提出了一种在传统双目位姿标定结果的基础上补偿双目标定矩阵改善精度的方法。介绍了摄像机模型、双目视觉测量模型和ICP算法的基本思想。用双目摄像机标定的外参数和相同的靶标坐标系获得双目视觉位姿矩阵,在此提出基于ICP算法获得两组点集的旋转平移矩阵补偿双目位姿矩阵的方法,以及相应的靶标角点坐标投影误差分析模型。双目摄像机采集9组5×7个角点的靶标标定图像,应用ICP算法补偿双目位姿矩阵,并采用误差模型对9组标定结果进行了分析,双目结构光标定改进实验结果表明,应用ICP算法补偿双目标定模型能显著地提高双目标定的精度。     

基于HALCON的板料成形网格应变测量系统的实现

随着计算机技术和信息技术的发展,视觉检测技术被引入到板料成形应变测量中。通过基于先进的图像处理软件HALCON,利用双目立体视觉原理,对摄像机进行标定,以获取摄像机的内外参数,并分析图像数量对标定准确度的影响。利用视差原理获得匹配点的三维坐标,根据圆形网格变形后的椭圆长短轴的长度变化,计算真实应变,并分析匹配误差对三维重建的影响,用OpenGL语言将应变可视化;在VC6.0下调用HALCON的算子库编程实现板料成形网格应变测量系统,实验证明该系统在速度、准确度方面能够达到测量要求。     

基于直线的射影不变性和极线约束标定摄像机参数

提出一种用图像对标定摄像机参数的新方法。传统自标定方法需要先由图像特征计算基础矩阵,该方法则是从图像中提取一组未知长度和长度比的平行直线,以图像中心坐标为主点初值,利用直线的射影不变性先求得摄像机焦距、旋转矩阵和平移向量的初值;再由图像对固有的极线约束来优化摄像机的参数,从而得到高精度的主点、焦距、旋转矩阵和平移向量。使用该方法对图像目标进行三维重建,仿真实验表明：该方法对场景要求较少,原理简单,方法易行,标定摄像机参数精度较高。     

基于点线对应的多摄像机全局位姿估计迭代算法

为了解决工业视觉场景中点和直线段共存情况下的多摄像机位姿估计问题,提出了基于点线特征对应的多摄像机全局位姿估计算法.以正交迭代算法为基础,使用多个固定在平台上的已标定摄像机间的几何约束,以全局方式处理多摄像机的重叠和非重叠视图的点和直线段特征,获得更加稳定精确的位姿估计.该算法先把所有摄像机数据进行统一表达,再把摄像机观测到的全部点特征的目标空间共线性误差和直线段特征的目标空间共面性误差之和作为误差函数,最后经数学推导得到使该误差函数最小化的迭代求解过程.实验结果验证了该算法的有效性以及优越性.     

基于点线对应的多摄像机全局位姿估计迭代算法

为了解决工业视觉场景中点和直线段共存情况下的多摄像机位姿估计问题,提出了基于点线特征对应的多摄像机全局位姿估计算法.以正交迭代算法为基础,使用多个固定在平台上的已标定摄像机间的几何约束,以全局方式处理多摄像机的重叠和非重叠视图的点和直线段特征,获得更加稳定精确的位姿估计.该算法先把所有摄像机数据进行统一表达,再把摄像机观测到的全部点特征的目标空间共线性误差和直线段特征的目标空间共面性误差之和作为误差函数,最后经数学推导得到使该误差函数最小化的迭代求解过程.实验结果验证了该算法的有效性以及优越性.     

基于三线结构光的相机平面标定方法研究

相机标定技术是结构光三维视觉测量的关键技术之一,结构光系统的相机标定的精度对三维测量的精度有很大影响。首先对三线结构光系统图的相机标定方法进行了分析,简单介绍了工业相机成像的几何模型及标定的原理;其次利用Harris角点检测方法提取特征点坐标,并选用了BP神经网络来校正工业相机的畸变模型,以提高标定算法的优化速度和标定精度;最后采用张正友的平面标定法对校正后的摄像机模型进行标定实验,由实验结果知,该方法具有一定的准确性和有效性,在一定误差范围内,基于神经网络畸变校正的张正友相机标定能够有效提高视觉检测的精度。     

Camera calibration approach using circle-square-combined target

Calibrating a small field camera is a challenging task because the traditional target with visible feature points that fit the limited space is difficult and costly to manufacture.We demonstrate a novel combined target used in camera calibration.The tangent points supplied by one circle located at the center of a square are used as invisible features,and the perspective projection invariance is proved.Both visible and invisible features extracted by the proposed feature extraction algorithm are used to solve the calibration. The target supplies a sufficient number of feature points to satisfy the requirements of calibration within a limited space.Experiments show that the approach can achieve high robustness and considerable accuracy. This approach has potential for computer vision applications particularly in small fields of view.     

一种微小型导引头视场下偏角的标定方法

描述了一种应用于微小型光学导引头视场下偏角标定的方法，通过获得导引头绕自身回转轴线转动时拍摄的固定靶标图像计算不同旋转角度时导引头中心视线落点和摄像机坐标系原点在世界坐标系下的坐标，以拟合获得的这2组点的分布建立空间几何关系进而得出下偏角。给出了3组实验数据，分别检验该方法的标定误差、实际所需的采样点数并给出了实际工作中的应用效果。实验表明该方法可以快速标定导引头系统的下偏角，其标定绝对误差小于0.10°。     

Using the camera pin-hole model restrictions to calibrate the lens distortion model

Camera calibration required the computation of camera pin-hole and lens distortion models. The lens distortion is estimated alone or together with the pin-hole model, by using some existing lens distortion non-metric or self-calibration methods. If both models are computed together, then the models are adjusted to training data, but not to real camera. This is because both pin-hole and lens distortion models are coupled. If they are computed separately, difficulties arise since calibration of lens distortion alone is an unstable process. To improve existing camera calibration methods, this paper proposes a metric calibration method to compute lens distortion separately from the pin-hole model. This method is solved under stable conditions, independently of the computed lens distortion model, since pin-hole and distortion models are computed separately. Images of a planar template are used. First, using distorted control points extracted from images, a set of undistorted points which fits in the pin-hole model are computed. Second, with distorted and undistorted control points, lens distortion is calibrated by using a metric calibration process.     

基于光学参考棒的立体视觉测量系统现场标定技术

为实现大空间复杂工件的准确测量,精确标定立体视觉系统变得越来越重要。为了克服传统立体摄像机标定过程繁复、户外实现困难的弱点,提出了一种基于光学参考棒的灵活、有效的立体视觉测量系统标定技术。参考棒水平和深度方向各有三个距离已知的红外LED作为特征点。通过在测量范围内的不同位置和方位移动光学参考棒,两像机同时捕获参考棒上特征点的图像。基于匹配的特征像点以及对极线约束,利用线性算法和Levenberg-Marquardt(LM)迭代算法快速地标定立体视觉测量系统。两像机之间平移量的比例因子由参考棒上特征点间的已知距离确定。参量标定过程中,自动地控制光强,优化曝光时间,使不同位置处光点图像的强度均一致,可以获得高的信噪比,提高标定精度。实验结果表明,该方法灵活、有效,在线标定能达到很高的精度,将现场标定过程应用到实际的大空间三维测量系统中,测量最大误差为0.18 mm。     

Camera calibration by using fringe patterns and 2D phase-difference pulse detection

Camera calibration is the most essential and usually the first step in computer vision applications. The results highly depend on the accuracy of the feature extraction. Conventional feature extraction methods suffer from perspective and lens distortion. In this paper, a novel feature extraction method is proposed by using fringe patterns groups as calibration target. Each group comprises six sinusoidal, in which, three are used for calculating vertical phases and the other three for horizontal phases. A three-step phase shift algorithm is used for wrapped phase calculation. Then, feature points are detected with a 2D phase-difference pulse detection method and refinement is done by simple interpolation. Finally, camera calibration is done by using these features as control points, experimental results indicated the proposed method accurate and robust.     

Out-of-focus color camera calibration with one normal-sized color-coded pattern

Conventional camera calibration methods always require well-focused target images for accurate feature detection. This requirement brings many challenges for long-range vision systems, due to the difficulty of fabricating an equivalent size target to the sensing field-of-view (FOV). This paper presents an out-of-focus color camera calibration method with one normal-sized color-coded pattern as the calibration target. The red, green and blue (RGB) channels of the color pattern are encoded with three phase-shift circular grating (PCG) arrays. The PCG centers used as feature points are extracted by ellipse fitting of the 2π-phase points. Experiments demonstrate that the proposed method can achieve accurate calibration results even under severe defocus.     

Building an accurate 3D model of a circular feature for robot vision

In this paper, an accurate 3D model analysis of a circular feature is built with error compensation for robot vision. We propose an efficient method of fitting ellipses to data points by minimizing the algebraic distance subject to the constraint that a conic should be an ellipse and solving the ellipse parameters through a direct ellipse fitting method by analysing the 3D geometrical representation in a perspective projection scheme, the 3D position of a circular feature with known radius can be obtained. A set of identical circles, machined on a calibration board whose centres were known, was calibrated with a camera and did the model analysis that our method developed. Experimental results show that our method is more accurate than other methods.