Efficiency of non-linear lens distortion models in biomechanical analysis of human movement

Accurate camera calibration is required in applications, which involve quantitative measurements such as biomechanical analysis of the human movement. Some information concerning the measuring of human movement can be obtained by using uncalibrated cameras, however lens calibration is necessary when metric information is needed. When large field of view is required, wide-angle lenses are useful but they show significant non-linear distortion in the image. The aim of this study was to compare the accuracy of distortion models, which are used in biomechanical analysis of the human movement. A new calibration pattern was designed and the image of the calibration pattern was captured by using a wide-angle lens. The distorted image of the pattern was automatically processed to determine the dot centroids in the image. Lens distortion was modelled accounting for three sources of distortion: radial, decentring and thin prism distortions. These three effects result in seven distortion coefficients (k₁, k₂, k₃, p₁, p₂, s₁, and s₂) in vertical and horizontal directions. Marzan and Karara (1975) [9] suggested 12, 14 and 16 parameters along with the Direct Linear Transformations (DLT) methods that comprise radial and decentring distortion models. In this study thin prism distortion [6] was also included to observe its effect. These non-linear distortion functions were minimised by using three different numerical methods, which were Least Squares, Levenberg–Marquard (LM) and Gauss–Newton (GM). In conclusion, the lens distortion models typically improved the accuracy. Applying the Non-linear Least Squares Optimisation Method (LM) showed the highest accuracy among the distortion models. While the average and the maximum error in the distorted image were 3.394% and 10.994%, respectively, they decreased to 1.591% and 3.524% in the application of DLT 18 method.     

基于共线特征点的摄像机镜头畸变校正

提出一种基于未知世界坐标共线特征点的未标定摄像机畸变校正算法以及基于预计算模板的快速实现方法。利用直线的透视投影不变性来估计镜头的畸变参数,不需要知道摄像机的内外部参数,也不需要知道任何空间特征点的世界坐标,简化了镜头畸变校正的过程。该方法的特点是需要的已知条件少,算法简单,易于实施,算法的计算速度快。试验结果表明,该方法能准确快速校正由摄像机镜头畸变引起的图像几何畸变,对机器视觉系统是可行的,预计算模板实现方法大大地降低了计算量,为镜头畸变校正实时应用奠定了基础。     

航空变焦距斜视成像几何畸变的自动校正

针对变焦距航空摄像机斜视成像产生的几何变形,提出一种同时校正斜视梯形失真和变焦距镜头非线性畸变的自动校正方法。根据直线透视投影不变性原理,利用单参数除式模型通过变步长优化搜索方法得到不同焦距对应的镜头畸变系数和畸变中心坐标;研究了焦距变化对畸变的影响规律,校正了镜头畸变使其满足针孔成像模型;引入飞机位置、姿态和摄像机视轴指向方位等因素,将航空图像重投影到地图坐标系中,对坐标变换后的像素亮度值进行重采样得到校正斜视变形和镜头畸变后的正射投影图像。对不同焦距和位置姿态下拍摄的地面靶标畸变图像和实际航空变焦距斜视图像进行了校正。结果表明,该方法能够有效准确地校正图像的几何变形,当飞行高度为2 500m时,在文中给定的位置姿态精度下的图像几何校正均方误差约为2m,较好地满足了后续图像拼接需求。该方法效率高,便于自动化实现,对提高图像拼接精度和实现目标精确定位与实时稳定跟踪具有重要意义。     

单层钎焊金刚石砂轮表面磨粒全场快速测量

为了快速获取单层钎焊金刚石砂轮表面全场磨粒的二维分布信息,提出了一种基于线阵相机的快速测量方法。对检测系统的原理、清晰无运动失真图像的获取、图像处理关键算法等进行研究。首先给出了检测系统的硬件构成与检测系统的工作原理；使用8方向模板的Sobel算子和标准件,完成相机对焦和运动失真图像矫正,然后对预处理的图像采用改进的阈值分割法Otsu算子对图像进行分割,最后实现了磨粒坐标及等效面积等二维特征量的提取。对磨粒有序排布的单层钎焊砂轮（粒度号35/40）进行测量实验，实验结果表明，系统从拍摄砂轮全场磨粒图像到获取磨粒坐标及等效面积所用的时间为77.4 s，可满足对单层钎焊砂轮的全场磨粒快速测量的要求。     

基于双平行平面相机模型的钢板尺寸视觉测量

提出基于双平行平面相机模型的视觉测量方法,用于测量生产线上运动钢板的尺寸。该方法采用数据驱动的方式计算像点在标定平面上投影点的世界坐标;采用k近邻(k-NN)方法生成目标在标定平面上的无畸变投影图像,并建立投影图像与世界坐标系的直接关联。提出了双平行平面模型下相机光心位置标定算法,利用线结构光进行板材厚度测量;在无畸变的投影图像上利用钢板边缘间的平行和垂直性进行钢板边缘特征提取,通过边缘直线的世界坐标方程求取长宽尺寸。最后,给出了针对大尺寸钢板测量的多相机测量系统框架。提出的方法为单目视觉测量方法,相比于其他方法具有现场安装简单和标定工作量小的特点。通过图像分辨率为640×480的相机对尺寸为80mm×50mm×15mm的标准铝块进行了测量,结果显示:厚度测量误差为0.1mm,长度和宽度的误差在0.2mm以内。实际应用中测量精度远高于加工精度,能够满足产品计量的要求。     

共面点的摄像机非线性畸变校正

采用传统的Tsai两步法进行摄像机标定时,标定精度会受一阶径向畸变模型的限制。本文提出了一种同时考虑摄像机镜头径向畸变和切向畸变的摄像机模型并研究了模型求解方法来提高畸变校正精度。考虑图像中心区域畸变较小,故用中心附近点列出线性方程组计算了摄像机的部分参数;建立了综合畸变模型,将摄像机参数代入模型计算畸变参数的初始值。由于焦距和平移分量在标定板与相机平面的距离深度变化不够时难以一次性准确标定,故将其代入综合畸变模型重新计算,并运用两步迭代法逐步逼近精确解。最后,对世界坐标系进行空间几何变换、透视变换和成像变换得到的重投影图像的像素坐标并与实际测得的像素坐标值进行比较,得到校正误差。结果表明,本文的畸变校正方法平均像素误差可以达到0.114 9pixel,优于Tsai校正方法的0.367 0pixel,且重复性较好。     

Accuracy evaluation of a smartphone-based technology for coastal monitoring

This paper introduces a smartphone-based technique for coastal monitoring and evaluates the accuracy of data such as images, three-dimensional coordinates, and attitude that can be acquired by such a technique. First, to determine intrinsic orientation (IO) parameters of a smartphone camera, a camera calibration was performed. The results were similar or slightly better than previous studies using a non-metric camera. 3-D coordinates provided by the assisted GPS (A-GPS) embedded in the smartphone showed lower accuracy. Attitudes calculated using an accelerometer and magnetometer showed 0.33–2.04° standard deviation, compared with ω, φ, and κ of extrinsic orientation (EO) parameters. Additionally, accuracy (RMSE 0.681 pixels) of smartphone image triangulation using ground control points (GCPs) was about equal to the RMSE of 0.404 pixels of a metric camera. Finally, ortho-rectified images of a test field were generated using DEM from terrestrial laser scanning and acquired images. Additionally, mapping of a shoreline was performed using the ortho-rectified images, and a profile of a representative cross-shore was composed. The results described the actual intertidal zone well. Given the results of the cross-shore profile and the horizontal and vertical accuracy evaluation of the extracted shoreline, this smartphone-based technique is considered appropriate for applications in coastal monitoring.     

单目摄像机线性模型可靠性分析与研究

要利用摄像机进行高精度的非接触式测量,通常在线性模型的基础上采用单目摄像机进行光学标定、成像、校正和测量,其中常用的针孔相机模型的可靠性及适用范围鲜有报道,为此对目前广泛采用的线性模型的可靠性进行分析与研究。根据单目摄像机成像特点,建立基于图像像素坐标与三维世界坐标的摄像机数学模型,利用自制模板进行视觉测量实验,并通过亚像素角点检测方法得到图像像素坐标,在不考虑光学畸变的情况下,对模板线段进行特征提取和线性度分析,并不断调整测量距离进行重复实验。实际针孔相机模型测量结果显示相对误差不超过1%,在该模型下的重复测量误差最大不超过0.3 mm,并进一步提出了该模型在不同区域和不同物距下存在的测量差异。实验结果显示,线性模型简单实用且在一定物距范围内具有良好的线性度。实验结果为针孔相机模型的建立提供了参考。     

A novel approach for the field calibration of line structured-light sensors

A new approach for the field calibration of line structured-light sensors is presented, which is fulfilled using a planar target and a raising block. The camera model is firstly established according to the mapping between the world coordinate frame and the computer image coordinate frame. Based on the rule of cross-ratio invariance the calibration points are extracted from the squares on the target, and the lens distortion coefficients are simultaneously acquired. The unknown parameters in the camera model are solved using these points. On the projection of the laser plane, two intersection lines are formed by the laser plane with the target and the raising block, respectively. The angle between the laser plane and the target is worked out by fitting plane using the two lines. Utilizing this angle, the three coordinates of a point in the laser plane can be represented by two of them. Thus the two coordinates can be directly obtained from the camera model. This method simplifies the calibration procedure of structured-light sensors, and facilitates online use. Experiment studies show that the calibrated sensor possesses good accuracy.     

Computer aided digitized microscope (CADIM) for nuclear emulsion analysis

The CADIM apparatus is a combination of : (a) a microscope equipped with a motorized stage and a drawing attachment; (b) high precision digitizers of X, Y and Z coordinates; (c) television camera, monitors and tape recorder; (d) a microcomputer and its peripherals. In addition to fast scanning and measurements in nuclear emulsion, it enables the performing of a direct visual comparison between the topology of an event seen in emulsion and its reconstructed topology from the data taken by associated electronic detectors. The use of television techniques allows image display and storing. This apparatus will be used in the first place for the CERN hybrid experiment WA71 searching for beauty particles.     

GPS双目摄像机标定及空间坐标重建

为减轻双目摄像机标定过程中对高精度靶标的依赖,实现摄像机参数的精确标定,并对空间坐标进行高精度重建,提出一种GPS双目摄像机标定及空间坐标重建方法,采用GPS代替2D或3D靶标进行双目摄像机标定。将GPS的位置在视场中任意移动,由被标摄像机拍摄多组含有GPS的图像,利用空间三维坐标与图像二维坐标间的映射关系,结合摄像机成像模型和双目摄像机标定原理,标定出双目摄像机参数,并对空间坐标进行精确重建。通过空间重建坐标与GPS实际测量值之间的相对距离误差,对重建精度进行检验。实验证明,该方法能够克服双目摄像机标定过程中对高精度靶标的依赖,空间重建坐标具有较高的精度,相对距离误差从1.56%减小到0.52%。     

基于双经度模型的鱼眼图像畸变矫正方法

针对鱼眼镜头拍摄图片桶形畸变大的问题,提出了一种基于双经度图像畸变矫正算法,以改善其视觉效果。首先,在未知镜头视角大小且鱼眼图像非圆形的情况下,利用鱼眼图像特征求得球面中心及其半径,并对半径大小做优化处理以减弱极点畸变;在此基础上采用双经度模型方法,通过正交投影策略将鱼眼图像映射到球面上,转化为球面横向经度和纵向经度坐标,从而投射为以横向、纵向双经度坐标为基础的正方形平面图像,实现图像的快速、精确变换。最后,利用2个实例验证了该文所提算法的有效性和可行性。     

智能三坐标测量机中零件位置自动识别系统

提出了一种基于机器视觉和CAD三维模型的零件自动定位方案，利用虚拟图像匹配的方法确定零件在工作台上的位置和方向。在CAD三维模型中，针对零件的可能放置方式通过透视成像的方法形成多幅虚拟图像。CCD产生的实际图像与各幅虚拟图像进行匹配。确定零件坐标系与机器坐标系的关系，达到定位目的。最后在三坐标测量机平台上实现了该方法，并给出了定位定向实验误差分析。     

Evaluation of three-dimensional surface roughness parameters based on digital image processing

With the development and application of optoelectric technology, laser technology and computer technology in mind, we have developed a method to evaluate three-dimensional surface roughness using surface profile information. This article proposes a three-dimensional measuring technique, which is used to survey components surface roughness, based on the digital image processing technology, and establishes athree-dimensional surface roughness evaluation system consisting of both hardware and software architecture. The hardware used in the present experiment is listed as follows: a stereomicroscope, a digital camera with special interface, a parallel light (by halogen lamp), an X, Y bidirectional laboratory bench and a computer. A computer-aided system (CAS), developed with Visual C⁺⁺ , is used in the image pretreatment and data processing analysis. In the experiment, image information gathered from the digital camera is pre-processed by median filtering, grayscale equalization and histogram conversion amplification. Then the data are analyzed by normalized cross-correlation and surface fitting techniques. Lastly, the correlation between m, σ, Sq, S _ku and Ra is discussed for different surface roughness specimens.     

基于眼模型的非球面眼底荧光相机的设计

运用非球面光学理论,设计了一款基于Gullstrand-Le Grand眼模型的便携式非球面眼底荧光相机。该相机根据滤光片选择原则,选用Nikon公司的B-2A滤光组合作为滤光系统进行光谱滤光;为避免角膜中心较大曲率引起的杂散光,采用环形光阑和共轴式照明相结合的照明方式实现眼底均匀照明;摄影系统中引入眼模型结构,综合校正了人眼及系统的像差,实现了全视场20×106pixel的高清晰成像。该摄影系统包含有2个非球面结构,使系统由9片镜结构简化为7片镜结构。实验结果表明,该相机具有较大的调节能力,对-12～+12m-1的人眼普遍适用,其视场角为30°,像面成像分辨率为120lp/mm,畸变值3.9%。非球面光学元件的使用,有效地简化了系统,减小了系统体积和重量,提高了成像质量。     

光学系统动态像点移动的坐标变换法

以宇宙飞船为载体的航天相机在轨工作过程中,由于受到空间动力学干扰源的作用,飞船、整个光学系统以及光学元件可能会产生振动,并由此引起曝光时间内像点的移动。这些振动必将造成光学系统成像质量的下降。本文探讨了振动过程中像点移动的分析方法--坐标变换法,并利用这种方法计算了航天相机主镜振动时的像点移动。     

Image distortion correction for micromanipulation system based on SLM microscopic vision

Stereo light microscope (SLM) simulates stereo imaging principle of human eyes. Microscopic vision system based on SLM has become an important visual tool for micro measurement, micromanipulation, and microinjection. We develop a micromanipulation system based on SLM and present an image distortion correction method. We mainly correct two kinds of image distortions: lateral and vertical distortion. Distortion correction consists of two steps. First, a linear fitting algorithm for each row or column of target points is developed, and the fitting errors are calculated. If the fitting errors are smaller than a given threshold, the linear fitting results are kept and used. Otherwise polynomial fitting procedure will be used. Second, the parallelism of straight lines is corrected. The results show that a line in world coordinate frame (WCF) is not necessarily a straight line in image coordinate frame (ICF), or two parallel lines in WCF may be not parallel in ICF. Distortion correction can restore the parallel and linear relationship. For distorted left and right images, the magnitude of distortion exceeds 6 pixels and 4 pixels in the horizontal direction, and 1.2 pixels and 1.7 pixels in the vertical direction, respectively. After corrected, for left and right image, distortion can be reduced to 0.8 pixels and 0.7 pixels in the horizontal direction, and 0.96 pixels and 1.3 pixels in the vertical direction, respectively. The results show that distortion parameters obtained from the proposed method can effectively correct distorted images. Microsc. Res. Tech. 79:162–177, 2016. © 2016 Wiley Periodicals, Inc.     

一种数码相机几何畸变的检测方法

针对数码相机普遍存在的几何畸变现象,提出了一种基于数字图像处理的畸变检测方法.在阐述数码相机畸变产生原因和检测原理的基础上,将待测数码相机拍摄目标靶板后的图片输入计算机,利用MatLab工具对图片进行数字图像处理,由像素间距获取实际像大小,由几何光学演算获得理想像大小,从而计算出相对畸变量.实验表明,该方法过程简单,易...     

Potential of functional image processing technique for the measurements of contact area and contact pressure of a radial ply tire in a soil bin testing facility

The exploration principally deals with a promising applicability of image processing method for contact area determination. A digital camera mounted on a single-wheel tester was utilized to take images of contact area at four levels of tire inflation pressure (i.e. 70, 100, 140, and 175 kPa) and six levels of wheel load (i.e. 0.75, 1.75, 2.75, 3.75, 4.75, and 5.75 kN). Contact pressure measurements, subsequently, were performed utilizing seven embedded load cells incorporated in an instrumented housing in longitudinal and perpendicular direction of traversing. It was observed that contact area is a polynomial (with order of two) function of wheel load while there was a reverse linear relation between contact area and inflation pressure. Furthermore, contact pressure has linear relation with vertical load and inflation pressure. Models for contact area and contact pressure were developed by multiple regression analysis while a classical predicting model was assessed using contact area results of image processing technique. The models indicated acceptable coefficient of determinations amounts of 0.974 and 0.953 for contact area and contact pressure models, respectively. However, applied image processing method denoted R²-value of 0.988 for contact pressure equation.     

标定模板与图像平面平行时的摄像机标定方法

在工业应用的机器视觉系统中,存在着较多的图像平面与标定模板平面平行的情况,针对这种情况,提出了一种新的简单而快速的标定方法。该方法仅需要一个取景器和一个平面标定模板,通过取景器可直接得到光轴中心坐标,同时也可计算出镜头的焦距;再利用平面标定模板上的特征点计算出摄像机模型的其他参数。实践证明,该方法无需预标定过程即可迅速、方便地对摄像机的内外参数进行标定,并且具有较好的精度。