Online self-calibration for mobile vision based on laser imaging and computer algorithms

A self-calibration technique for mobile three-dimensional vision is presented. This technique determines the vision parameters during the vision task based on computer algorithms and image processing. The three-dimensional vision is performed by a Bezier network based on laser line projection. This network provides the data to perform the online self-calibration when the vision system is modified. Here, the changes of the extrinsic and intrinsic parameters are determined. The structure of the network is performed by the line shifting provided by the surface depth. From this structure, the data for the initial calibration and online self-calibration are deduced. In this manner, the calibrated references and physical measurements are avoided to perform the online self-calibration. Therefore, calibration limitations caused by online modifications are overcome to perform the mobile vision. Thus, the proposed self-calibration improves the accuracy and performance of the mobile vision. It is because online data of calibrated references are not passed to the vision system. This procedure represents a contribution in the field of the online recalibration, which is performed based on calibrated references. To elucidate this contribution, an evaluation is performed based on the self-calibration methods, which are reported in the recent years. Also, the time processing is described.     

Laser imaging and approximation networks for calibration of three-dimensional vision

An automatic calibration technique for three-dimensional vision is presented. This technique determines the vision parameters by means of approximation networks and image processing of a laser line. Also, the network performs the three-dimensional vision based on laser scanning and a mobile camera. The network is built based on the behavior of a laser line and the camera position. Here, an electromechanical device moves the camera during the vision task and the network provides an automatic re-calibration. Thus, the vision is carried out via mobile camera and extrinsic parameters are not passed to the system. Therefore, the accuracy and the performance of the vision procedure are improved. It is because the physical measurements of references are avoided. Thus, a contribution for the calibration of lighting methods is achieved. Additionally to it, the mobile setup provides a high sensitivity and avoids occlusion. The proposed technique is tested with real objects and its results are evaluated based on laser projection methods. To carry it out, a root mean square of error is computed using a contact method as reference. Also, the time processing is described.     

基于点阵编码的三维主动视觉标定

提出一种基于位错点阵编码三维成像系统的标定方法.首先在物空间建立三维数据基准,然后通过基准传递的概念标定摄像子系统,再通过建立摄像子系统坐标系与投影子系统坐标系之间的约束关系,将标定后的摄像子系统的准确度传递到投影子系统坐标系.摄像子系统坐标系与投影子系统都具有标定的准确度之后,可以根据位错点阵编码三维成像技术的解码算法获得深度图像空间坐标的计算值,然后将其与物体空间的三维标定数据基准进行比较,建立目标函数为误差平方和最小的非线性优化方程.通过迭代求解这个优化方程,最终获得三维系统的结构参量.实验结果表明,经过三维标定的位错点阵编码三维成像系统,对300×300×80 mm3的测量体积内,可以获得X方向的标准差为0.29mm和Y方向的标准差为0.24mm,Z方向的标准差为0.29mm的测量准确度.     

基于光学测棒的立体视觉坐标测量系统的研究

提出了一种基丁测量与校准功能合一的光学测棒的立体视觉坐标测最系统.采用光学测棒作为成像目标,通过任意放置的两台摄像机获取测棒上的发光特征点的图像实现被测物体三维坐标的测量,同时利用测量数据定期对两台摄像机外部方位参数进行校准.深入研究了两台摄像机内部参数和外部方位参数校准过程中的校准件和校准算法的设计,以及系统测量建模等关键技术,提出了相应的解决方案,减小了摄像机内外参数校准及测量模型对测量结果的影响,提高系统的测量精度.实验结果表明.该系统的最大测量误差为0.11 mm.     

Axial stereo-photogrammetry for 360° measurement on tubular samples

This work presents a stereo-photogrammetry (SP) based procedure to perform whole-body measurements on tubular samples. Such a system is designed for future applications to the study of vascular wall mechanics. The use of a concave conical mirror surrounding the specimen makes it possible to capture the reflected 360° surface with a single snapshot moving neither cameras nor object. Then, according to 3-D computer vision principles, a stereo camera system retrieves control points depth information from image-pairs of the investigated surface. An axial-SP arrangement is selected since is more suitable for this specific application than the more popular lateral-stereo model.In this paper, particular emphasis is given to a formulation taking into account even small camera misalignments. A calibration process based on optimization concepts is used together with a feature-based matching algorithm to efficiently find correspondence between highly distorted images reflected by the conical mirror.Both theoretical and experimental analyses on calibrated samples demonstrated feasibility and accuracy of the proposed procedure.     

Calibration for stereo vision system based on phase matching and bundle adjustment algorithm

Calibration for stereo vision system plays an important role in the field of machine vision applications. The existing accurate calibration methods are usually carried out by capturing a high-accuracy calibration target with the same size as the measurement view. In in-situ 3D measurement and in large field of view measurement, the extrinsic parameters of the system usually need to be calibrated in real-time. Furthermore, the large high-accuracy calibration target in the field is a big challenge for manufacturing. Therefore, an accurate and rapid calibration method in the in-situ measurement is needed. In this paper, a novel calibration method for stereo vision system is proposed based on phase-based matching method and the bundle adjustment algorithm. As the camera is usually mechanically locked once adjusted appropriately after calibrated in lab, the intrinsic parameters are usually stable. We emphasize on the extrinsic parameters calibration in the measurement field. Firstly, the matching method based on heterodyne multi-frequency phase-shifting technique is applied to find thousands of pairs of corresponding points between images of two cameras. The large amount of pairs of corresponding points can help improve the accuracy of the calibration. Then the method of bundle adjustment in photogrammetry is used to optimize the extrinsic parameters and the 3D coordinates of the measured objects. Finally, the quantity traceability is carried out to transform the optimized extrinsic parameters from the 3D metric coordinate system into Euclid coordinate system to obtain the ultimate optimal extrinsic parameters. Experiment results show that the procedure of calibration takes less than 3 s. And, based on the stereo vision system calibrated by the proposed method, the measurement RMS (Root Mean Square) error can reach 0.025 mm when measuring the calibrated gauge with nominal length of 999.576 mm.     

A trinocular vision system for close-range position sensing

A machine vision system for obtaining three-dimensional information from a single camera is presented. Stereo and orthogonal views of the test object are imaged onto a CCD (charge coupled device) camera by using a tri-split lens in conjunction with a mirror arrangement. The calibration procedure involves a direct linear transformation resulting in an overdetermined set of equations that are solved to obtain the three-dimensional coordinate position. The position information obtained from the stereo views is used to locate the object in the orthogonal view, thus improving the accuracy of the position sensor. The relative position of a cylinder-head from a reference point, which is of importance in assembly line applications, was measured using the vision system within an accuracy of 0.3% of full scale value.     

Constructing feature points for calibrating a structured light vision sensor by viewing a plane from unknown orientations

To calibrate a structured light vision sensor, it is necessary to obtain at least four non-collinear feature points that fall on the light stripe plane. We propose a novel method to construct non-collinear feature points used for calibrating a structured light vision sensor with a planar calibration object. After the planar calibration object is moved freely in the range of measuring of the structured light vision sensor at least twice, all the local world coordinates of the feature points falling on the light stripe plane can be readily obtained in site. The global world coordinates of the non-collinear feature points in the local world coordinate frame can be computed through the three-dimensional (3D) camera coordinate frame. A planar calibration object is designed according to the proposed approach to provide accurate feature points. The experiments conducted on a real structured light vision sensor that consists of a camera and a single-light-stripe-plane laser projector reveal that the proposed approach has high accuracy and is practical in the vision inspection applications. The proposed approach greatly reduces the cost of the calibration equipment and simplifies the calibrating procedure. It advances structured light vision inspection one step from laboratory environments to real world use.     

Simple laser vision sensor calibration for surface profiling applications

Due to the relatively large structures in the Oil and Gas industry, original equipment manufacturers (OEMs) have been implementing custom-designed laser vision sensor (LVS) surface profiling systems as part of quality control in their manufacturing processes. The rough manufacturing environment and the continuous movement and misalignment of these custom-designed tools adversely affect the accuracy of laser-based vision surface profiling applications. Accordingly, Oil and Gas businesses have been raising the demand from the OEMs to implement practical and robust LVS calibration techniques prior to running any visual inspections. This effort introduces an LVS calibration technique representing a simplified version of two known calibration techniques, which are commonly implemented to obtain a calibrated LVS system for surface profiling applications. Both calibration techniques are implemented virtually and experimentally to scan simulated and three-dimensional (3D) printed features of known profiles, respectively. Scanned data is transformed from the camera frame to points in the world coordinate system and compared with the input profiles to validate the introduced calibration technique capability against the more complex approach and preliminarily assess the measurement technique for weld profiling applications. Moreover, the sensitivity to stand-off distances is analyzed to illustrate the practicality of the presented technique.     

Robust geometric,phase and colour structured light projection system calibration

This paper introduces a new comprehensive procedure for both geometric and colour calibration of structured light system. In order to perform both geometric and colour calibration procedure, a new calibration artifact is proposed. The intrinsic and extrinsic parameters of projector and camera are estimated by using an extended pinhole camera model with a tangential and radial distortion. Camera image plane coordinates are obtained by extracting features from images of a calibration artifact. Projector image plane coordinates are calculated on the basis of continuous phase maps obtained from a fringe pattern phase reconstruction procedure. In order to stereo calibrate camera-projector system, pairs of corresponding image plane points are calculated with subpixel accuracy. In addition, one of three pattern views is used in colour calibration. RGB values of a colour field pattern detected by camera and their reference values are compared. This comparison leads to derivation of a colour transformation matrix. The performance of the proposed method is tested by measuring plane, sphere and distance reference. Also 360 degrees complex object 3D model from a set of measurements is obtained. Residual mean errors for all tests performed are calculated.     

Innovative calibration technique for fringe projection based 3D scanner

A new approach to 3D scanners calibration, based on Fourier Transform Profilometry method, is proposed and verified by experiments. Commonly used calibration techniques require that a calibration plane is placed in different and known positions in the scanner measuring volume; an optimization procedure iteratively change the value of some calibration parameters in order to minimize the discrepancy between the height distributions of the measured planes and the reference ones. The main limit of this type of calibration techniques is the necessity to guarantee a high accuracy in the placement of the planes used as calibration references. The innovation of the method proposed in this paper is the complete removal of any device for the accurate plane placement by assigning the task of the plane positions estimation to the camera, which is part of the scanner. The well known camera calibration algorithm proposed by Zhang permits at the same time to calibrate the camera and to estimate the positions and orientations of the plane used for the calibration itself. The knowledge of the plane position allows to use them as a reference of the scanner calibration. The obtained results show that this method provides accuracy values for the scanner parameters estimation comparable with other calibration techniques, but the advantage is that no particular device is needed.     

“网格式平行线法”在螺纹参数测量中的应用

在计算机视觉理论基础上发展起来的视觉检测(vision inspection)技术具有非接触、速度快、精度适中、可实现在线等优点，已广泛地应用于工业产品的在线检测。在计算机视觉检测技术中CCD摄像机是一个最关键的器件，其参数是否准确决定了检测的精度。所以，摄像机标定是视觉检测技术中最基本的也是最重要的一步。在比较其他标定方法的基础上，为了解决传统标定方法对螺纹图像测量系统所带来的一系列问题，采用了一种新的图像测量系统的标定方法——网格式平行线标定方法，该方法运用了CCD亚像素细分技术及调焦技术，可直接得到纵横2个方向的像素，经过理论分析及实验结果均表明，采用该方法具有标定简单、精度高、重复性好等优点，是一种较好的螺纹图像测量系统标定方法。     

CBERS-02B星HR相机内方位元的在轨标定方法

遥感相机内方位元素的在轨标定对遥感图像的定位和测量具有重要意义.利用线阵推扫传感器构像模型,提出了一种对CBERS-02B星HR相机内方位元素进行在轨标定的方法.该方法以内方位元素和姿态角为未知参数,建立地面控制点和相应像点的共线方程组,通过解算共线方程组获得内方位元素.实验证明,用该方法对内方位元素进行在轨标定,具有...     

A novel optimization method of camera parameters used for vision measurement

Camera calibration plays an important role in the field of machine vision applications. During the process of camera calibration, nonlinear optimization technique is crucial to obtain the best performance of camera parameters. Currently, the existing optimization method aims at minimizing the distance error between the detected image point and the calculated back-projected image point, based on 2D image pixels coordinate. However, the vision measurement process is conducted in 3D space while the optimization method generally adopted is carried out in 2D image plane. Moreover, the error criterion with respect to optimization and measurement is different. In other words, the equal pixel distance error in 2D image plane leads to diverse 3D metric distance error at different position before the camera. All the reasons mentioned above will cause accuracy decrease for 3D vision measurement. To solve the problem, a novel optimization method of camera parameters used for vision measurement is proposed. The presented method is devoted to minimizing the metric distance error between the calculated point and the real point in 3D measurement coordinate system. Comparatively, the initial camera parameters acquired through linear calibration are optimized through two different methods: one is the conventional method and the other is the novel method presented by this paper. Also, the calibration accuracy and measurement accuracy of the parameters obtained by the two methods are thoroughly analyzed and the choice of a suitable accuracy evaluation method is discussed. Simulative and real experiments to estimate the performance of the proposed method on test data are reported, and the results show that the proposed 3D optimization method is quite efficient to improve measurement accuracy compared with traditional method. It can meet the practical requirement of high precision in 3D vision metrology engineering.     

Research on camera calibration using new optimization strategy

An important task for stereo vision is camera calibration, whose goal is to obtain the intrinsic and extrinsic parameters of each camera. This paper proposes a new accurate calibration method with multilevel process of camera parameters. In order to improve the calibration accuracy, a sub-pixel corner detection method is presented. We start with several views of a planar calibration to obtain some intrinsic camera parameters and to build an accurate model with lens distortion on a planar calibration target. Flexibly making use of geometry imaging theory, our algorithm obtains all the parameters through logical organization of solving order, accordingly avoids obtaining possible local optimized problem when solving the non-linear equation, gets over the relativity influence of every unknown parameters of traditional calibration way, and makes the error distributed among the constraint relation of parameters. Experiments with real images are carried out to verify the image correction effect and numerical robustness of our results. Compared with classical calibration techniques, that use expensive equipment and complicated mathematical computation, the proposed technique, which was verified by experiment, achieves high accuracy and reliable parameters.     

基于圆结构光视觉传感器的标定特征点获取方法

管道作为工业生产重要的传输手段其内表面腐蚀程度和瑕疵的精确检测对于保证安全生产具有重要意义。针对管道内表面圆结构光视觉检测,提出了一种基于共面参照物获取圆结构光视觉传感器标定特征点的新方法。该方法设计了圆结构光平面靶标,基于交比不变原理,以摄像机三维坐标系为中介,将多个局部世界坐标系下的标定特征点统一到全局世界坐标系中,得到位于圆结构光曲面上的非共线标定特征点的三维世界坐标。该方法降低了标定设备的成本,简化了结构光视觉传感器的标定过程。标定实验精度达到0.340 mm,标定结果表明,该方法切实可行。     

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。     

基于特征面片的复杂型面轮廓视觉测量方法

提出一种基于特征面片的复杂型面轮廓三维视觉测量方法.该方法包括多模式局部特征检测与匹配、特征面片集初始化、膨胀、滤波等重构过程,其输入是已标定的立体图像,输出是覆盖表面的可视稠密面片集.该方法不需要目标和场景的任何初始边界信息,能够自动检测和抛弃外部点和缺失部分;不需要对邻域特征进行任何平滑处理,仅需通过局部单演特征一...     

基于构建初始测量网络的相机内部参数校准

提出了一种基于构建初始测量网络的相机内部参数校准方法,有效解决了二维平面靶标在深度方向信息的不足以及三维靶标空间的限制等问题。通过对靶标板进行初始成像,按照测量网络的构建原则建立初始测量网络,经过后方交会求解外方位、前方交会求解空间三维靶标点坐标,最后由光束平差优化求解相机内部参数。利用标定后内部参数求解空间点坐标,实验结果表明,采用构建初始测量网的误差平均值为0.0794,优于平面靶标和立体靶标标定的-0.2443和-0.1916。同时校准所用时间也明显小于虚拟立体校准,即该方法具有快速、精确和方便等优点,满足大空间视觉测量中相机内部参数现场校准的要求。     

CCD camera automatic calibration technology and ellipse recognition algorithm

A novel two-dimensional (2D) pattern used in camera calibration is presented. With one feature circle located at the center, an array of circles is photo-etched on this pattern. An ellipse recognition algorithm is proposed to implement the acquisition of interest calibration points without human intervention. According to the circle arrangement of the pattern, the relation between three-dimensional (3D) and 2D coordinates of these points can be established automatically and accurately. These calibration points are computed for intrinsic parameters calibration of charge-coupled device (CCD) camera with Tsai method. A series of experiments have shown that the algorithm is robust and reliable with the calibration error less than 0.4 pixel. This new calibration pattern and ellipse recognition algorithm can be widely used in computer vision.