基于交比不变量的摄像机标定方法

为了解决摄像机标定的菲线性优化过程中解的不稳定性问题,针对具有镜头径向畸变的摄像机模型,提出了一种实用的摄像机线性标定方法.该方法利用透视投影的交比不变性原理标定镜头的畸变系数,基于摄像机成像的单应性矩阵及内部参数的基本约束,线性标定出其他摄像机参数.实验结果表明,该方法标定模板制作简单,速度快,具有一定的鲁棒性,可以满足工业机器视觉的精度要求.     

基于计算机视觉的线性摄相机标定技术研究

通过建立基于一阶径向畸变的摄像机数学模型,提出了一种线性求解摄像机参数的标定方法,对摄像机模型中内部参数和外部参数进行逐步求解,且全部过程采用线性方法求解,从而改变了传统的摄像机标定依赖于非线性优化的缺点,避免了非线性优化的不稳定性.实验证明此方法具有较高的标定精度和较强的实用性.     

改进的工程机器人立体视觉标定方法

针对摄像机线性标定准确程度不高与非线性标定计算复杂等问题,以己知的三维信息为基础推导二维像素点对信息,提出了基于神经网络的三目立体视觉摄像机标定方法,构建了改进的BP神经网络模型,对比分析了两种摄像机标定方法的像素点对均方根误差。结果表明:采用改进的BP神经网络能够避免对摄像机进行非线性建模,有利于提高标定精度,增加系统的灵活性,更具有实际意义。     

基于图像畸变矫正的摄像机标定方法

本文提出了一种新的摄像机标定方法,首先对含有畸变的二维图像进行矫正,再应用矫正后的图像对摄像机进行线性模型标定。给出了标定方法的理论依据和求解过程,并将其与线性标定方法和两步法进行了对比实验。实验结果表明:该方法无需预标定过程,即可迅速、方便地对摄像机进行标定,并且具有较好的精度。     

基于Harris-张正友平面标定法的摄像机标定算法

为了提高双目摄像机标定参数的精确性,以张氏标定方法为基础,首先采用Harris算法对黑白棋盘表格进行内角点提取,其次将内角点像素点坐标精确到亚像素精度,最后利用投影线性关系求出双目摄像机的内外等参数,并与cv Find Chessboard Corners()函数提取内角点标定方法进行对比,证明了该方法准确、有效,提高了黑白棋盘表格内角点检测的准确性和标定精度,可以应用于双目视觉系统中.     

一种摄像机标定系统的实现

摄像机标定系统中的参数直接影响摄像机三维重建效果的逼真度。本文以张正友两步标定法为基础,设计制作标定模板,并采用针孔摄像机模型对摄像机标定参数的求解过程进行了优化;结合OpenCV库,在Matlab开发环境下实现了摄像机标定系统。实验结果表明,该方法提高了摄像机的标定精度。     

双目视觉立体标定方法的改进

为了提高双目立体视觉系统中立体标定的精度,提出了一种利用正方形标定域中相互正交方向的消失点进行立体标定的新方法.首先摄像机从不同的方向拍摄平面标定板,在图像平面上确定一个正方形投影区域,再根据射影几何模型中的交比性质,计算出正方形中相互正交方向上的两对消失点;然后利用消失点与光心连线的射影几何性质,线性求解摄像机的全部内外参数;最后利用全局优化方法进行参数的优化.这一方法继承了张正友标定方法畸变模型的使用,实验利用MATLAB通过左右摄像机拍摄的两组图像对方法进行验证,实验结果表明,该方法的立体标定精度较高,且旋转矩阵R和平移向量T均符合实际要求;由于实验中正方形投影区域的大小及具体的位置关系是随机的,使其具有较好的实用性.     

基于平面靶分割区间LS-SVM摄像机标定的研究

摄像机标定是视觉系统精确的前提.利用最小二乘支持向量机实现摄像机的标定不需确定摄像机具体的内部参数和外部参数.由于镜头的畸变主要由径向畸变引起,根据摄像机畸变的特点,对畸变区域进行划分,提出一种基于分割区间最小二乘支持向量机的摄像机标定法,对不同的畸变区域进行单独处理.该方法同BP神经网络和基本最小二乘支持向量机标定预测结果对比表明,基于分割区间最小二乘支持向量机的摄像机标定法速度快,实时性好,能有效提高标定精度.     

多个不同精度摄像机的协同标定

针对成像性能不同的多个摄像机,提出了一种协同标定算法。算法主要由预处理数据和协同标定组成。预处理数据利用二分法快速找出错误特征点,并删除;协同标定使得一个精度高或观察点好的摄像机能帮助精度低或观察点不好的摄像机消除误差大的数据,提高标定精度,并对协同标定作了误差分析。实验表明,所提出的多个摄像协同标定算法适应性强,当数据组足够时可以明显消除外部实验环境给标定算法带来的误差。     

基于移动靶机器视觉测量系统的标定

论述了机器视觉测量系统中摄像机方位元素标定的基本原理、物镜系统畸变参数以及系统中其他误差处理的基本思想.在三坐标机上测出标定块上各孔心在标定块板面的二维坐标通过移动标定块产生标定点列的另一维坐标经角度变换采用优化方法求取摄像机的各参数并经验后补偿校正提高了精度     

Calibration of Camera with Large Field-of-View Based on Flexible Planar Target

For high precision calibration of camera with large field-of-view, massive calibration points will be needed if traditional methods are selected, which makes the calibration complex and time-consuming. In order to solve this problem, a calibration method based on flexible planar target is proposed. In this method, distortion factor is firstly acquired by the invariance of cross ratio, and existing feature points are adjusted with the distortion factor. Then, a large number of points that will be used for the calibration are constructed with the adjusted feature points. Simultaneously, Tsai method is modified so as to reduce the complexity of calibration, which makes the process linear. The simulation and real experiments show that the method proposed in this paper is simple, linear, accurate and robust, and the precision of this method is close to that of Tsai method using abundant points. The method can satisfy the requirement of high precision calibration for camera with large field-of-view.     

一种基于参考平面的直线距离测量方法

根据射影几何中的灭点原理提出了一种基于参考平面的直线距离测量方法。在已知标定相机内部参数的前提下,利用放置在参考平面（如地平面）上,已知尺寸的平面矩形作为标定物,建立相机与参考平面间的空间关系。在相机位置和焦距不变的情况下,拍摄一张数字图像,可以利用基于参考平面的直线距离测量方法计算参考平面中任意两个特征点间的直线距离。提出的测量原理和方法具有定标物和测量步骤简单、测量精度高的特点,在工程测量领域具有广泛的应用价值。     

基于单CCD尺寸检测的摄像机标定方法

为提高机械零件尺寸检测的精度,针对Tsai两步标定法中初始参数不精确的问题,提出了基于直线投影约束的三步标定方法.该方法首先利用直线投影约束条件求解出图像中心点和畸变参数的初始值,然后结合径向排列约束条件求解超定方程组得出全部外部参数和余下内部参数值,最后对全局参数进行非线性优化.为验证提出方法的效果,通过建立基于图像模板的世界坐标系实现,即首先利用图像像素坐标求取摄像机坐标,再将摄像机坐标进行归一化,利用外参数矩阵就可以求解出图像点的世界坐标.结果表明,该方法可以有效提高摄像机标定的精度,像素误差可达0.126 5像素点,单目视觉尺寸检测中任意两点距离误差小于0.41%.     

基于交比不变性的相机镜头畸变标定研究

相机的标定对于整个摄影测量系统的精度起着至关重要的作用.主要研究了一种较为有效、易行的标定相机畸变参数的方法.标定物简单且容易获得,相机无须固定,可手持拍摄.利用交比不变的性质,拟合全视场的无畸变的虚拟网格,得到图像点的偏差,从而计算出畸变系数.最后通过比较畸变纠正前后空间点距离的测量,表明此方法可较为明显的修正镜头畸变,且简单易行.     

双线阵CCD交汇测量中的自动标定方法

针对CCD交汇测量系统中的初始标定困难、繁琐、重复误差等缺陷,提出了一种基于测量模型的自动标定方法.根据交汇测量原理,建立了测量系统的数学模型,并推导出自标定的模型.在相机焦距、倾角、光心坐标已知条件下,通过分析标定中各参量的函数关系,可利用线阵相机像元确定测量位置,精度可达到4 mm.     

考虑一阶径向畸变的摄像机参数标定

针对GRB-400机器人视觉系统,建立完整的摄像机一阶径向畸变模型。首先用GRB系统的Matrox图像采集卡采集标定点坐标,然后利用径向排列约束方法 （简称RAC）建立线性方程组求解摄像机的径向畸变参数。基于此畸变参数实现图像校正,以提高对目标物体的定位精度,从而提高机器人视觉系统对摄像机标定参数变化的鲁棒性。实验结果表明：该方法有较高的标定精度。     

基于单模板的二维场景重建方法

针对现有场景重建算法具有计算复杂度高且鲁棒性差的缺点,本文采用一种基于映射变换的黑箱标定方法,根据图像坐标与世界坐标的线性映射模型,利用矩阵方程获取摄像机参数。与传统方法相比,该方法仅需一幅模板图像,降低了特征点与模板个数,简化了计算。实验过程中,利用图像降噪、改进的Harris算子角点提取等预处理过程快速提取特征点,实现摄像机参数标定的映射变换;建立了误差评价函数评价本文方法与Tsai标定算法。实验结果表明,该方法具有稳定性强、计算精度高、算法简捷的特点。     

Research on a New-style Visual Sensor for Measurement

First,the constitution of traditional visual sensor is presented.The linear camera model is introduced and the transform matrix between the image coordinate system and the world coordinate system is established.The basic principle of camera calibration is expatiated based on the linear camera model. On the basis of a detailed analysis of camera model,a new-style visual sensor for measurement is advanced.It can realize the real time control of the zoom of camera lens by step motor according to the size of objects.Moreover,re-calibration could be avoided and the transform matrix can be acquired by calculating,which can greatly simplify camera calibration process and save the time. Clearer images are gained,so the measurement system precision could be greatly improved.The basic structure of the visual sensor zoom is introduced,including the constitute mode and the movement rule of the fixed former part,zoom part,compensatory part and the fixed latter port.The realization method of zoom controlled by step motor is introduced. Finally,the constitution of the new-style visual sensor is introduced,including hardware and software.The hardware system is composed by manual zoom,CCD camera,image card,gearing,step motor,step motor driver and computer.The realization of software is introduced,including the composed module of software and the workflow of measurement system in the form of structured block diagram.     

Research on a New-style Visual Sensor for Measurement

First, the constitution of traditional visual sensor is presented. The linear camera model is introduced and the transform matrix between the image coordinate system and the world coordinate system is established. The basic principle of camera calibration is expatiated based on the linear camera model.On the basis of a detailed analysis of camera model, a new-style visual sensor for measurement is advanced. It can realize the real time control of the zoom of camera lens by step motor according to the size of objects.Moreover, re-calibration could be avoided and the transform matrix can be acquired by calculating, which can greatly simplify camera calibration process and save the time. Clearer images are gained, so the measurement system precision could be greatly improved. The basic structure of the visual sensor zoom is introduced, including the constitute mode and the movement rule of the fixed former part, zoom part, compensatory part and the fixed latter part. The realization method of zoom controlled by step motor is introduced.Finally, the constitution of the new-style visual sensor is introduced, including hardware and software. The hardware system is composed by manual zoom, CCD camera, image card, gearing, step motor, step motor driver and computer. The realization of software is introduced, including the composed module of software and the workflow of measurement system in the form of structured block diagram.     

Non-Metric CCD Camera Calibration Algorithm in a Digital Photogrammetry System

Camera calibration is a critical process in photogrammetry and a necessary step to acquire 3D information from a 2D image. In this paper, a flexible approach for CCD camera calibration using 2D direct linear transformation （DLT） and bundle adjustment is proposed. The proposed approach assumes that the camera interior orientation elements are known, and addresses a new closed form solution in planar object space based on homogenous coordinate representation and matrix factorization. Homogeneous coordinate representation offers a direct matrix correspondence between the parameters of the 2D DLT and the collinearity equation. The matrix factorization starts by recovering the elements of the rotation matrix and then solving for the camera position with the collinearity equation. Camera calibration with high precision is addressed by bundle adjustment using the initial values of the camera orientation elements. The results show that the calibration precision of principal point and focal length is about 0.2 and 0.3 pixels respectivelv, which can meet the requirements of close-range photogrammetry with high accuracy.