CN103411535B - A variable weight for a return light reflected ghost marker point positioning - Google Patents

A variable weight for a return light reflected ghost marker point positioning Download PDF

Info

Publication number
CN103411535B
CN103411535B CN201310340841.2A CN201310340841A CN103411535B CN 103411535 B CN103411535 B CN 103411535B CN 201310340841 A CN201310340841 A CN 201310340841A CN 103411535 B CN103411535 B CN 103411535B
Authority
CN
China
Prior art keywords
point
weight
spot
value
method
Prior art date
Application number
CN201310340841.2A
Other languages
Chinese (zh)
Other versions
CN103411535A (en
Inventor
董明利
庄炜
孙鹏
王君
燕必希
祝连庆
娄小平
Original Assignee
北京信息科技大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京信息科技大学 filed Critical 北京信息科技大学
Priority to CN201310340841.2A priority Critical patent/CN103411535B/en
Publication of CN103411535A publication Critical patent/CN103411535A/en
Application granted granted Critical
Publication of CN103411535B publication Critical patent/CN103411535B/en

Links

Abstract

本发明提供了一种针对回光反射标志的可变权重像点定位方法,所述方法包括如下步骤:a)设置多个物方标志点;b)在不同站位对所述多个物方标志点进行拍摄,获取成像光斑上的多个位置点处的灰度值;c)获取和背景噪声有关的第二权重因子β;d)获取与测量场环境变量相关的第一权重因子α;e)根据所获取的第一权重因子α和第二权重因子β,计算所测量椭圆形像点的点中心坐标(x0,y0)。 The present invention provides a variable weightings for the return light reflected ghost point positioning marker, the method comprising the steps of: a) providing a plurality of landmark object side; b) at said plurality of different stations of the object side landmark shooting to acquire gradation value at a plurality of points on the imaging position of the light spot; c) obtaining a background noise and a second weighting factor related β; d) obtaining a first weight associated with the measurement field environment variable weight factor [alpha]; e) a first weight obtained weight factors α and the second weighting factor β, the point position of the elliptical image of the measured point is calculated (x0, y0). 根据本发明的针对回光反射标志的中心定位方法较以往方法更为通用,可以适用于不同的现场环境,尽可能地消除了现场环境造成的影响,提高了回光反射标志点中心的定位精度。 According to the Centers positioning method for reflecting light back marker of the invention is more versatile than the conventional method can be applied to different site environment as much as possible to eliminate the influence of the field of the environment, improve the positioning accuracy of light reflected back to the center point mark .

Description

一种针对回光反射标志的可变权重像点定位方法 A variable weight for a return light reflected ghost marker point positioning

技术领域 FIELD

[0001] 本发明涉及摄影测量技术领域。 [0001] The present invention relates to the field of photography measurements. 具体而言,本发明涉及一种针对回光反射标志(Retro-ReflectiveTargets,简称RRT)的可变权重像点定位方法。 In particular, the present invention relates to a back light for the reflective markers (Retro-ReflectiveTargets, referred to as RRT) variable weighted image point positioning method.

背景技术 Background technique

[0002] 摄影测量利用摄影获得被测物的二维图像,求出被测物的三维信息。 [0002] Photogrammetry using a two-dimensional photographic image of the measured object, obtaining three-dimensional information measured object. 20世纪80 年代中期,近景摄影测量在工业应用获得成功,其利用大幅面网格相机,多画面收敛计算, 数字图像扫描、处理、比较等技术,提供了高达1:500,000的相对精度,对大尺寸物体(如, 直径>10米)的测量可超过经炜仪的精度。 The mid-1980s, close-range photogrammetry success in industrial applications, which use large-format camera grid, multi-screen convergence of computing, digital image scanning, processing, comparison and other technology to provide up to 1: 500,000 of relative accuracy, large size of the object (e.g., a diameter of> 10 m) measurement accuracy may exceed by Wei instrument. 90年代中期,结合了数字传感器(CXD&C0MS) 技术后,近景摄影测量被称为离线摄影测量系统。 The mid-1990s, a combination of digital sensors (CXD & C0MS) after art, close range photogrammetry Photogrammetry is called off-line system. 通过高分辨率单反相机,回光反射标志点,以及像素定位的技术,其能够提供很高的精度水平。 High resolution SLR, landmark reflection return light, and a pixel location technology, capable of providing a high level of accuracy. 其典型的测量精度为1:100,〇〇〇到1:200,000相对精度(1倍标准差),前者对应10米目标尺寸下达到0. 1mm的绝对精度,而一般用长度测量的绝对精度来表征三维重建的相对精度,一般比点定位精度低2到3倍(例如,对2m的物体约0• 05mm)〇 Typical measurement accuracy of 1: 100, 〇〇〇 to 1: 200,000 relative accuracy (1 times the standard deviation), the former corresponding to 0. 1mm to achieve the target size of the absolute accuracy of 10 meters, the general accuracy absolute length measurement to Characterization of the relative accuracy of three-dimensional reconstruction, the positioning accuracy is generally lower than the point 2 to 3 times (e.g., from about 2m to the object 0 • 05mm) square

[0003] 回光反射标志是近年来国内外高精度的工业摄影测量中广泛使用的一种人工标志,其采用特殊的具有高折射率的玻璃微珠材料结构。 [0003] The return light reflective markers abroad in recent years an artificial marker precision industrial photogrammetry widely used in which a special glass having a high refractive index material structure microbeads. 在特定位置使用光源(例如闪光灯) 照射时,RRT能将入射光按照原来的路径反射到光源处,如图1所示。 A light source (e.g., a flash) is irradiated at a specific position, the RRT capable of reflecting incident light according to the original path to the light source, as shown in FIG. 其具有极高的反射率, 一般是普通标志反射率的几百倍甚至上千倍,以达到目标点与背景"分离"的准二值图像, 有助于目标点的精确定位和特征点的提取与识别。 Quasi binary image having high reflectance, the reflectance is generally normal sign of several hundred or even thousands of times to reach a target point and the background "isolated" help pinpoint the feature point and the target point extraction and recognition.

[0004] 在获得了准二值图像后,即可对RRT进行定位。 [0004] After obtaining the quasi-binary image, it can be positioned on RRT. 通常RRT是一个有一定大小的圆形,其圆心位置即该RRT所代表的坐标位置。 RRT is a generally circular certain size, i.e., the coordinate position of its center position represented by the RRT. 为提取这一位置,针对不同情况有不同的方法。 To extract this position, there are different ways for different situations. 在相机感光元件为胶片的时代,通常是将胶片置于高倍率放大镜下,得到一个放大后的RRT图像,然后通过几何方法找到RRT的中心位置。 In the camera sensor is the era of the film, the film is typically placed under high magnification magnifier, to obtain an enlarged image RRT and RRT find the center position of geometric methods.

[0005] 随着现代科技与制造技术的发展,尤其是以CCD和C0MS为重要代表的固体图像传感器技术的研制与飞速发展,采用数字像机来获取数字影像已经越来越普及,基本上已经取代了传统的胶片式像机。 [0005] With the development of modern science and technology and manufacturing technology, particularly in research and the rapid development of solid-state CCD image sensor technology and C0MS as an important representative of the digital camera to obtain digital images has become increasingly popular, basically replacing the traditional film type camera. 数字相机不像胶片相机那样,可以将传感元件取出放大来对RRT 进行定位,而由于CCD与C0MS传感器发展水平的限制,其空间分辨率远不如胶片的高。 As a digital camera, unlike a film camera, the sensing element may be removed to locate the enlarged the RRT, but due to limitations C0MS CCD sensor and level of development, which is far less high spatial resolution film. 圆形回光反射标志经透镜成像以后一般为椭圆形,成像光斑的示意图如图2所示。 Circular return light reflected by the mark after the imaging lens is generally elliptical, a schematic view of the imaging spot shown in FIG.

[0006] 受光学系统和图像传感器制造工艺的限制,图像的分辨率相对精度不超过1:10, 000,为达到更高的相对精度,需要对回光反射点进行亚像素级别的精确定位,即需要对成像椭圆光斑的中心进行定位。 [0006] limited by the optical system and the image sensor manufacturing process, the relative accuracy of the resolution of the image does not exceed 10, 000, to achieve relatively higher accuracy, the need for reflection return light spot for precise positioning of the sub-pixel level, the need for the center of the imaging elliptical spot positioning. 为了达到对椭圆中心的高精度定位,通常有两种方法。 In order to achieve precision positioning of the center of the ellipse, there are generally two methods. 一类方法是求取RRT边缘,然后拟合出椭圆方程,从而求得RRT椭圆点中心;另一类方法是基于RRT的灰度信息,通过质心法来求得RRT椭圆点中心。 Obtaining a class of methods RRT edge, and ellipse fitting equation to obtain the center point of the ellipse RRT; Another method is based on the RRT gray information to be obtained through the center point of the ellipse RRT centroid method. 其中边缘拟合法对边缘点提取的精度要求很高,在标志点大小较小的时,会因为边缘提取误差导致椭圆点中心受到严重影响。 Wherein the edge fitting of the edge points extracted high precision, at smaller point sizes flag, because errors due to edge extraction ellipse center point severely affected. 而基于灰度的质心法对边缘点提取的精度不敏感,在不同尺寸的标志点下提取精度变化不大,其缺点是容易受到灰度分布不均匀的影响。 Based on the gradation centroid of the edge point extraction accuracy is not sensitive to variations in the accuracy not extracted mark point different size, the drawback is susceptible to non-uniform intensity distribution.

[0007] 常用的亚像素定位算法有拟合法,数字相关法,形心法。 [0007] The common subpixel location fitting algorithm, digital correlation method, Centroid. 使用拟合法的前提是目标特性满足已知或假定的函数形式,所以一般较难获得很高的亚像素精度。 Fitting using the premise that meet certain characteristics known or assumed functional form, it is generally difficult to obtain a high sub-pixel accuracy. 数字相关法具有原理简单、适应性强和精度高等优点,但在对旋转目标或旋转没有约束的目标定位中很少使用相关的方法,因为这使模板的选择难以实现,而在近景摄影测量中,RRT总是不可避免地带有旋转,因此相关法不能用于RRT点中心定位。 Digital correlation method has simple, adaptable, and high precision, but rarely used in a related method for targeting rotating or rotating the target is not binding, since it is difficult to realize the selection of a template, in the close range photogrammetry , RRT always inevitably rotated, so the relevant law can not be used RRT point centering. 灰度重心法、带阈值的灰度重心法、平方加权法等是基于形心法形成的亚像素算法,这些算法的优点在于充分利用了RRT中每一点的灰度值,通常可获得比形心法更高的亚像素精度,精度在l/20px到1/50像素左右。 Barycenter method, barycenter method with the threshold value, the weighted square method or the like is formed based on sub-pixel Centroid algorithm, these algorithms advantage that full use of the gray value of each point in the RRT, shape than the generally available Heart higher subpixel accuracy, precision about l / 20px to 1/50 pixel. 但是根据不同文献的实验结果显示,这些方法的点中心提取精度优劣并不是绝对的,而是在不同的场合下,获得不同的点中心提取精度。 However, the experimental results show the different documents, the center point of the merits of these methods of extraction accuracy is not absolute, but in different situations, different points to obtain the center of the extraction accuracy. 因为这些方法受到不同现场环境的影响,从而表现出不一致的性能水平,因此在环境变化的情况下,使用起来非常不便。 Because these methods are subject to different field environment, thus showing inconsistent performance levels, so in the case of environmental change, it is very inconvenient.

[0008] 因此,需要一种更为通用的针对回光反射标志的中心定位方法,可以适用于不同的现场环境,尽可能地消除现场环境造成的影响,以提高回光反射标志点中心的定位精度。 [0008] Thus, a more general method for locating the center of the back light reflective markers can be applied to different site environment, eliminate the impact of on-site environmental impacts as much as possible, in order to improve the positioning of light reflected back to the center point mark accuracy.

发明内容 SUMMARY

[0009] 本发明的目的在于提供一种针对回光反射标志定位的变量加权质心法,以提高点中心的定位精度。 [0009] The object of the present invention is to provide a weighted centroid for variable positioning mark reflected return light to improve the accuracy of the positioning of the center point.

[0010] 根据本发明的一个方面,提供了一种针对回光反射标志的可变权重像点定位方法,所述方法包括如下步骤:a)设置多个物方标志点;b)在不同站位对所述多个物方标志点进行拍摄,获取成像光斑上的多个位置点处的灰度值;c)获取和背景噪声有关的第二权重因子0 ;d)获取与测量场环境变量相关的第一权重因子a;e)根据所获取的第一权重因子a和第二权重因子0,基于下式计算所测量椭圆形像点的点中心坐标(X(l,ytl), [0010] In accordance with one aspect of the present invention, there is provided a variable weight for a return light reflected ghost point positioning marker, the method comprising the steps of: a) providing a plurality of landmarks on the object side; at different stations b) the plurality of bits landmark photographing object side, the gradation value acquired at multiple locations on the spot imaging points; c) obtaining a background noise and a second weighting factor related to the right 0; d) acquiring field measurement environment variables associated with a first weighting factor; E) according to the acquired first weight and a second weight factor a weighting factor of zero, based on the point position of the elliptical image points measured from the following equation (X (l, ytl),

Figure CN103411535BD00051

[0013] 其中xJPy』为成像光斑上的特定位置点的座标(xi,y』),g(Xi,y』)为(Xi,y』)处成像光斑的灰度值,[Wi,%]为成像光斑的宽度范围,[h,h2]为成像光斑的高度范围。 [0013] wherein xJPy "is a specific coordinate position of a point on the imaging spot (xi, y"), g (Xi, y ") is (Xi, y") at a gray value image spot, [Wi,% ] the width of the imaging spot, [h,] the imaging spot height range h2.

[0014] 优选地,所述物方标志点为利用定向反光材料制作的圆形RRT标志。 [0014] Preferably, the object side for the use of a circular mark point RRT flag of retroreflective material.

[0015] 优选地,所述RRT标志的直径范围在3_12mm。 [0015] Preferably, the diameter of said marker in the RRT 3_12mm.

[0016] 优选地,所述RRT标志的直径为6mm。 [0016] Preferably, the diameter of the RRT flag to 6mm.

[0017] 优选地,所述物方标志点均匀地分布在被测场中。 [0017] Preferably, the object side marker points uniformly distributed in the measured field.

[0018] 优选地,所述第二权重因子0为所拍摄图像的平均灰度值。 [0018] Preferably, the second weight factor of zero for the average gray value of the captured image.

[0019] 优选地,所述第二权重因子0为0。 [0019] Preferably, the second weighting factor 0 is 0.

[0020]优选地,[Wi, w2]和IX,h2]取值范围略大于成像光斑的宽度%和高度h。 [0020] Preferably, [Wi, w2], and IX, h2] in the range slightly larger than the width of the imaging spot% and a height h. .

[0021] 优选地,所述步骤d)中第一权重因子a的最优值的计算方法如下:dl)测量求出所拍摄的图片上各个成像点的坐标,提取出求得的二维点中心坐标(XiCI,yi(l)作为真值;d2) 使用包含所述公式的算法求出图片上各点坐标(Xi,yi)作为测量值;d3)求出所述测量值与所述真值的残差的均值;d4)取最小均值对应的a为最优值。 The method of calculating the optimal value [0021] Preferably, said step d) a first weight factor as follows: dl) measured to obtain coordinates of each point on the imaging picture taken, extracted two-dimensional points obtained center coordinates (XiCI, yi (l) as the true value; D2) algorithm using the formula comprising obtaining coordinates of each point (Xi, yi) on the image as the measurement value; D3) obtains the measured value and the true the mean value of the residuals; D4) takes a minimum mean corresponding to an optimal value.

[0022] 优选地,所述第一权重因子a的取值为整数。 [0022] Preferably, the first weight factor is a value of an integer.

[0023] 根据本发明的针对回光反射标志的中心定位方法较以往方法更为通用,可以适用于不同的现场环境,尽可能地消除了现场环境造成的影响,提高了回光反射标志点中心的定位精度。 [0023] According to the Centers for positioning method back to the light reflective markers of the present invention is more versatile than conventional methods can be applied to different site environment as much as possible to eliminate the influence of the field of the environment, improve the reflection of light back to the center point mark positioning accuracy.

[0024] 应当理解,前述大体的描述和后续详尽的描述均为示例性说明和解释,并不应当用作对本发明所要求保护内容的限制。 [0024] It should be understood that the description and the following detailed description are exemplary of the foregoing general description and explanation and should not be used to protect the content of the present invention as claimed is restricted.

附图说明 BRIEF DESCRIPTION

[0025] 结合附图详细描述了本发明的上述和其他方面,附图中: [0025] described in detail in conjunction with the drawings The above and other aspects of the present invention, the drawings:

[0026] 图1示出了在特定位置使用光源照射时RRT反射入射光的光路示意图; [0026] FIG. 1 shows a schematic view of an optical path of incident light reflected at RRT specific position using the light source;

[0027] 图2示出了圆形回光反射标志经透镜成像后的成像光斑的示意图; [0027] FIG. 2 shows a schematic view of the imaging spots by the imaging lens circular return light reflective markers;

[0028] 图3示出了根据本发明的可变权重像点定位方法的流程图; [0028] FIG. 3 shows a flowchart ghost point positioning method according to the present invention, the variable weights;

[0029] 图4示出了验证本发明方法的实验用控制场示意图; [0029] FIG. 4 shows an experimental verification method of the present invention is a schematic diagram of a control field;

[0030] 图5示出了利用图4的实验用控制场拍摄的像点照片; [0030] FIG. 5 shows a picture dot experiment of Fig. 4 taken with the control field;

[0031] 图6示出了图像平面中权重系数a与二维坐标残差的关系曲线。 [0031] FIG. 6 shows an image plane in weight versus weight coefficient and a residual dimensional coordinates.

具体实施方式 Detailed ways

[0032] 本发明提出了一种针对回光反射标志的可变权重像点定位方法。 [0032] The present invention provides a variable weight for a return light reflected ghost marker point positioning method. 通过结合测量场的相关环境变量,将测量场的相关因素,例如角度、光照和标志点的分布,以及拍摄的背景噪声等设定为可以变化的权重因子从而综合进行考虑,提出了一种具有可变权重的计算回光反射标志中心位置的定位方法。 By combining the measurement field of environmental variables related to the measurement field factors, for example the angle distribution, light and landmarks, and the background noise is set to be shot weight factors may be varied so as to be considered comprehensive weight, proposes a positioning mark reflected return light of variable weights method of calculating the center position of the counterweight. 根据本发明的方法可以基于环境变量设定相应的权重, 因此适用于不同的测量环境,从而获得最优的权重系数,实验证明可以达到比现有技术更高的定位精度。 The method of the present invention may be set based on the environment variable corresponding weights, and therefore for different measurement environment, so as to obtain the optimum weighting coefficient, it proved to achieve a higher positioning accuracy than the prior art.

[0033] 图3示出了根据本发明的可变权重像点定位方法的流程图。 [0033] FIG. 3 shows a flowchart ghost point positioning method according to the present invention is variable weights.

[0034] 在步骤301,设置多个物方标志点。 [0034] In step 301, a plurality of landmark object side. 物方标志点为利用定向反光材料制作的圆形RRT标志,RRT标志的直径范围在3-12_,优选约为6_。 Landmark object side to use circular RRT flag retroreflective material having a diameter in the range RRT flag 3-12_, preferably about 6_. 优选地,物方标志点应尽量均匀地分布在被测场中。 Preferably, the object side marker points should be evenly distributed in the measured field.

[0035] 在步骤302,在不同站位对待测的所设置的多个物方标志点进行拍摄,获取成像光斑上的多个位置点(Xi,h)处的灰度值。 [0035] In step 302, a plurality of treatment measured object side mark point provided at different shooting stations, acquiring a plurality of location points (Xi, h) on the gray value at the spot imaging. 站位是指拍摄用的装置,例如相机拍摄图片时所处的位置,站位可以通过拍摄装置在全局坐标系的坐标(x,y,z)和拍摄装置的光轴指向对应的三个旋转角(0,《,k)来表示。 Station refers to a device used for imaging, for example, the position where the camera images, three stations may point to the coordinate axis corresponding to the rotation in the global coordinate system (x, y, z) and the photographing apparatus by imaging means angle (0, ", k) to represent. 拍摄装置站位的坐标分布应尽量均匀分布在以测对象中心为球心的球锥面上,锥角在75°~105°之间,一般取90°,光轴应尽量指向球心。 Imaging station coordinate distribution apparatus should be evenly distributed over the center of the sphere to the center of the target object tapered ball, a cone angle between 75 ° ~ 105 °, and generally 90 °, the optical axis center of the sphere should be directed.

[0036] 在步骤303,获取摄影装置的参数,即获取和背景噪声有关的第二权重因子0。 [0036] In step 303, a parameter acquiring imaging device, i.e., the background noise and obtaining a second weight related to the weight factor of zero. 通常可以使用所拍摄图像的平均灰度值作为第二权重因子0。 The average gradation value may be used generally captured image as a second weighting factor of zero. 在某些场合中,例如当RRT尺寸很大且亮度均匀的情况下,也可以忽略0。 In some instances, such as when the large size and uniform brightness RRT case, 0 may be ignored.

[0037] 在步骤304,获取与测量场环境变量相关的第一权重因子a,令(x。,yQ)为求得的所测量椭圆形像点的点中心坐标,(Xo,%)与第一权重因子《和第二权重因子0之间的关系推导如下: [0037] In step 304, obtaining a first weight associated with the measurement field environment variable weight factors a, so (x., YQ) of the point position of the elliptical image of the measured point obtained, (Xo,%) and the second a weighting factor "and the relationship between the second weighting factor of 0 is derived as follows:

[0038] 首先,令 [0038] First of all, make

Figure CN103411535BD00071

[0041] 其中,F(Xi,y」,kpk2,…kn)是与成像光斑上的特定位置点(Xi,y」)处灰度值相关的函数。 [0041] where, F (Xi, y ", kpk2, ... kn) and the specific location is the point (Xi, y ') on the spot imaging-related function of the gradation value. ki,k2,…kn为函数的系数,求取的范围为xf[w^wJyjG[h^hj,其中成像光斑的宽度范围,[hptg为成像光斑的高度范围。 ki, k2, ... kn is a coefficient function of obtaining the range xf [w ^ wJyjG [h ^ hj, wherein the width of the imaging spot, [hptg height range of the imaging spot. 优选地,[Wl,w2]和[hptg取值范围略大于成像光斑的宽度%和高度,一般取光斑宽度绝对值,S卩丨Wli2l=w#4,光斑高度绝对值,即丨丨=hQ+4 (4的单位为像素)。 Preferably, [Wl, w2], and [hptg% range slightly greater than the width and height of the image spot, generally the absolute value of the spot width, S Jie Shu Wli2l = w # 4, the absolute value of the height of the light spot, i.e. Shushu = hQ +4 (4 units of pixels).

[0042] 令(Xi,y」)处灰度值为g(Xi,y」),则可将⑴式中的F(Xi,y」,kpk2, •••!〇写作多项式形式,即 [0042] order (Xi, y ') at a gradation value g (Xi, y "), the formula may be the ⑴ F (Xi, y", kpk2, •••! Writing square polynomial, i.e.,

[0043] [0043]

Figure CN103411535BD00072

[0044] 对式(2)右侧展开后取两阶时,有 When the right take two stage Expanded [0044] formula (2), there is

[0045] F(x"y』,k"k2,…kn) •g(x"y』)2+k2 •g(x"y』)i+k3 (3) [0045] F (x "y", k "k2, ... kn) • g (x" y ") 2 + k2 • g (x" y ") i + k3 (3)

[0046] 变形得, [0046] The modification too,

[0047] F(xi;y』,k"k2,…kn) =1(g(xi;y』)+K2) 2+K3 ⑷ [0047] F (xi; y ", k" k2, ... kn) = 1 (g (xi; y ") + K2) 2 + K3 ⑷

[0048] 其中 [0048] in which

Figure CN103411535BD00073

[0049] 由(1)式可知,&取值不影响结果,故取1。 [0049] apparent from equation (1), does not affect the value of & result, it takes 1. 又从实际测量经验中得到1(2和1( 3对点中心定位的影响基本相同,因此为简化算法可将其合并,于是有 And from the actual measurement obtained in substantially the same experiences an influence (2 and 1 (3-point centrally located, so as to simplify the algorithm may be combined, so there

[0050] F(x"y』,k"k2,…kn) = (g(x"yj)-0 )a (5) [0050] F (x "y", k "k2, ... kn) = (g (x" yj) -0) a (5)

[0051] 其中a为多项式最高项n-1,|3 =K2,aGN+, |3彡-g(Xi,yj,即 [0051] wherein A is a maximum polynomial term n-1, | 3 = K2, aGN +, | 3 San -g (Xi, yj, i.e.

Figure CN103411535BD00081

[0054]在步骤305,根据所获取的第一权重因子a和第二权重因子0,基于式(6)计算所测量椭圆形像点的点中心坐标(x〇,y〇),即完成了像点的定位。 [0054] In step 305, according to the acquired first weight and a second weight factor a weighting factor of zero, based on Equation (6) computes an elliptic spot image center coordinates of the measured points (x〇, y〇), to complete the as anchor points.

[0055] 根据本发明的第一权重因子a反映了与测量场环境有关的各种变量,针对不同的环境场可以选用不同的a值,从而获得精度最高的定位方法。 [0055] According to the present invention, the first weight factor a weight and reflects the measurement field variables relating to the environment, may use different values ​​for a different ambient field so as to obtain the highest precision positioning method. 第一权重因子a的最优值的计算可以采用如下方式。 Calculating the optimal value of a first weight factor can be used in the following manner.

[0056] 可以测量求出所拍摄的图片上各个成像点的坐标(例如利用VSTARS软件),提取出求得的二维点中心坐标(XiCI,yi(l)作为真值;使用包含变权重模型(即上述的公式(6))的算法程序求出图片上各点坐标(Xi,yi)作为测量值;以测量值与真值的残差的均值;K作为评价标准,因为aeQ,所以残差均值S应随着a连续变化,且必定在特定a处取得最小值, 该a值即最优权重。 [0056] can be measured image coordinates of each point on the captured image is determined (e.g., using software VSTARS), extracts the center coordinates obtained two-dimensional point (XiCI, yi (l) as the true value; containing variable weight model (i.e., the above-mentioned equation (6)) algorithm program obtains coordinates of each point (Xi, yi) on the image as the measurement value; to measure the mean value of the residuals from the true value; K as an evaluation criterion, because AEq, so residues shall mean the difference S with a continuous change, and must be made at a minimum at a specific, i.e., the optimal weight value a.

[0057] 根据本发明的式(6)不仅可以包含环境变化的因素,即由a和0的取值来反映不同的测量场情况,还可以兼容现有的定位计算方法。 [0057] The factors of formula (6) according to the present invention may include not only changes in the environment, that is, from a value of 0 and to reflect the different circumstances of the measurement field, also compatible with the conventional method for positioning calculation. 例如,当a=l,0=0时,式(6)即为传统的灰度质心法,即, For example, when a = is the conventional gray-level centroid l, 0 = 0, the formula (6), i.e.,

[0058] [0058]

Figure CN103411535BD00082

[0059] 其中f(Xi,y」)是坐标为(Xi,y」)处点的灰度值,求取的范围为XiG[Wl,W2],yjG[hi,!^],(xQ,yQ)是求得的点中心坐标。 [0059] where f (Xi, y ') are coordinates (Xi, y ") at the point of gradation values, ascertained range XiG [Wl, W2], yjG [hi,! ^], (XQ, yQ) is obtained center point coordinates.

[0060] 当a=1,f3=th(th表示阈值)时,式(6)即为带阈值的灰度质心法,即, [0060] When a = 1, f3 = th (th represents a threshold value), the formula (6) is the centroid, with a gradation threshold value, i.e.,

[0061] [0061]

Figure CN103411535BD00083

[0062]其中f(Xi,yj)是坐标为(Xi,yj)处点的灰度值,th是阈值常数。 [0062] where f (Xi, yj) is the coordinates (Xi, yj) at the point of gradation values, th is the threshold constant.

[0063] 当a=2, 0 =0时。 [0063] When 2, 0 = 0 a =. 式(6)即为灰度平方加权质心法,即 Of formula (6) is the squared weighted centroid gradation, i.e.,

[0064] [0064]

Figure CN103411535BD00091

[0065] 其中f(Xi,y』)是坐标为(Xi,y』)处点的灰度值,求取的范围为XiG[Wl,w2], [ kh],这里对像素灰度值做了平方处理。 [0065] where f (Xi, y ') are coordinates (Xi, y ") at the point of gradation values, ascertained range XiG [Wl, w2], [kh], where the pixel grayscale values ​​do a square deal.

[0066] 由此可见,根据本发明的定位方法可以根据环境的变化适应不同的情况,当RRT 尺寸很大,亮度均匀的情况下,可以忽略0,令0=0,而取a=l;当拍摄得到的图像具有一定背景高斯噪声的情况下,可以设定0为一定的阈值;而当RRT尺寸不大,像素弥散斑影响不能忽略的情况下,可以通过设定a的大小来反映像素弥散斑的影响,例如取a=2或a=3 等等。 [0066] Thus, according to the positioning method of the present invention can be adapted according to changes in the environment are different, when the RRT large size, brightness uniform, 0 can be ignored, so that 0 = 0, and take a = l; in the case where an image photographed with a certain background Gaussian noise, 0 may be set to a certain threshold; RRT when the small size, the pixel blur spot influence can not be ignored, the pixel may be reflected by setting a size Effect of diffuse plaques, for example, take a = 2 or a = 3 and so on.

[0067] 以上方法中,a取值均为整数,因此对应不同的测量环境。 [0067] The above method, a is an integer of value are, thus corresponding to different measurement environments. 更优选地,在更为复杂多变的实际的测量环境下,a取值也可以不为整数。 More preferably, in the more complex environment of the actual measurement, the value A may not be an integer. 可以令aGQ。 You can make aGQ. 对不同的测量环境, 权重《的最优值不是固定值,而是随着测量场的相关环境变量决定,例如角度,光照,标志点的分布等。 Different measurement environments, the weights "optimum value is not a fixed value but as the measurement field of environmental variables related decisions, for example the angle, the illumination distribution of the landmarks and the like. which is

Figure CN103411535BD00092

[0069] 其中,〇£〇,0彡1(\4),1^,1^,1^为测量场的相关环境变量。 [0069] wherein, £ billion square, San 1 0 (\ 4), 1 ^, 1 ^, 1 ^ of the measurement field relating environment variables. 式(1〇)即反映了权重因子可变的像点定位方法。 Formula (1〇), that reflects the image point positioning a variable weighting factor.

[0070]实骀结果 [0070] Results solid forworn

[0071] 1•数据采集 [0071] 1 • Data Acquisition

[0072] 对本发明的可变权重定位方法进行实验验证。 [0072] The experimental verification of variable weights relocation method of the present invention. 控制场的示意图如图4所示,其大小为4m*3m*l. 5m。 Diagram of the control field shown in Figure 4, having a size of 4m * 3m * l. 5m. 在图4所示的控制场内布置若干个RRT,如图4所示是均匀分布的。 RRT arranged in a plurality of control field shown in FIG. 4, FIG. 4 is uniformly distributed. 图4 只是示意性地示出了一部分布点。 FIG 4 schematically shows only a portion of the distribution. 使用美国GSI公司研制的工业数字近景摄影三坐标测量系统V-STARSE4X进行拍摄。 GSI using the US company developed industrial digital close-range photogrammetry coordinate measuring system V-STARSE4X shoot. 拍摄装置采用尼康D2Xs,1200万像素,分辨率为4288X2848, 像素大小为6ym,使用环形闪光灯进行曝光。 Photographing means using the Nikon D2Xs, 1200 million pixels, a resolution of 4288X2848, pixel size 6ym, using a ring flash exposure. 处理软件为V-STARS4. 3,其点中心定位精度可达l/50px。 Processing software V-STARS4. 3, which is the center point of the positioning accuracy up to l / 50px.

[0073] 用尼康D2Xs相机在不同站位拍摄多张照片,共计67张。 [0073] to take multiple shots at different stations with a Nikon D2Xs camera, a total of 67. 拍摄得到的图片局部如图5所示。 Partial images photographed in Figure 5.

[0074] 2 •数据处理 [0074] 2 • Data Processing

[0075] 在图4所示的控制场环境下,由于使用了高质量的摄影器材,可控制背景噪声基本为0,则此时0=0。 [0075] In the environment of the control field shown in Figure 4, since the quality of photographic equipment, the background noise can be controlled to substantially 0, 0 = 0 at this time. 这样(6)式可化简为 Such (6) can be simplified to

[0076] [0076]

Figure CN103411535BD00101

[0077] 因为最优权重系数a由测量场的一些相关因素决定,例如角度、光照、标志点的分布,即a=F(k1;k2,…,kn) [0077] Since the optimum weighting factor is determined by a number of factors related to the measurement field, e.g. distribution angle, illumination, landmarks, i.e. a = F (k1; k2, ..., kn)

[0078] 其中的函数关系F目前还不明确,故使用比较的方法寻找当前测试场环境下的最优权重系数a。 [0078] where the function F is not yet clear, it is using the comparative method to find the optimal weight in the current market environment test weight coefficient a. 在求最优权重时,使用以下方法: When seeking the optimal weights using the following method:

[0079] 使用VSTARS软件求出图片上各点坐标,提取出VSTARS求得的二维点中心坐标(xiCI,yiCI)作为真值;使用包含变权重模型的算法程序,求出图片上各点坐标(Xi, yi)作为测量值;以测量值与真值的残差的均值S作为评价标准,因为aEQ,所以残差均值S应随着a连续变化。 [0079] software using VSTARS obtains coordinates of each point on the image, the center coordinates of the extracted two-dimensional point VSTARS obtained (xiCI, yiCI) as the true value; algorithm program containing heavy variable weight model, obtains the coordinates of each point on the image (Xi, yi) as the measurement value; to the measured value and the true value of the mean of the residuals of S as an evaluation criterion, because AEq, the mean residual S so as to be continuously changed a. 且必定在特定a处取得最小值,该a值即最优权重。 And it must be made at a minimum at a specific, i.e., the optimum value of a weight. 数据处理结果如图6所示。 The data processing results shown in Figure 6.

[0080] 由图6可见,残差均值X随权重系数a在[1.0,2.0]区间内变化,并在a约为1. 6时取得最小值。 [0080] FIG 6 can be seen from the residual weight coefficients with X mean a change in the [1.0, 2.0] interval, and obtaining the minimum value is about 1. 6 a. 与a=1. 〇的传统质心法和a=2. 0的灰度平方加权质心法相比较,提高了至少25%。 And a =. 1. Billion traditional centroid and a = gray squares weighted centroid wears 2.0 compared increased by at least 25%. 由上述方法求得的最优权重a,在取最优权重时,评价指标均好于传统方法, 说明使用变权重灰度质心法提取点中心的精度要好于传统方法。 Determined by the method described above optimal weight a, while taking optimal weight, are better than the conventional evaluation method, the accuracy of instructions of variable weight centroid extraction point gradation center is better than conventional methods.

[0081] 通过上述实验验证了根据本发明的可变权重定位方法,其点中心提取精度要优于传统质心法和灰度平方加权质心法,通过计算残差的标准差作为评价定位方法的标准,其中的残差是指根据本发明的方法求得点中心坐标Coord(u,v)与坐标真值CoordCKi^,V(l)间的差值的模IICoord-CoordOl|。 [0081] Verify the relocation method according to the present invention, variable weights, the center point extraction accuracy of better than traditional centroid and gray squares weighted centroid by calculating the standard deviation of the residuals as an evaluation standard positioning methods by the above experiment wherein means residuals obtained center coordinates point according to the invention is Coord (u, v) coordinate true value CoordCKi ^, the difference between the V (l) molding IICoord-CoordOl |. 实验结果表面,残差的标准差至少能提高25%以上。 Experimental results of surface, the residual standard deviation of at least 25% or more increase. 因此, 使用根据变权重模型后,点中心求取精度相对传统的方法有提高。 Thus, the relative accuracy using traditional methods has been improved in accordance with the model variable heavy weights, the center point is obtained. 按照变权重模型,最优权重a不是固定值,是而随不同的测量场而变化。 Model variable weight according to the weight, the optimal weight is not a fixed value, but with a different measurement field varies. 具体的最优权重系数a由测量场的一些相关因素决定,例如角度、光照、标志点的分布。 DETAILED optimum weighting factor is determined by a number of factors related to the measurement field, e.g. distribution angle, illumination, landmarks.

[0082] 应当理解的是,在不背离后附权利要求所述的本发明精神的前提下,本发明可有多种组合、修正、改变和替换方案。 [0082] It should be appreciated that in the appended claims without departing from the spirit of the premise of the present invention, the present invention may have a variety of combinations, modifications, changes, and alternatives. 以上所述仅为本发明中的具体实施方式,但本发明的保护范围并不局限于此,任何熟悉该技术的人在本发明所揭露的技术范围内,可理解想到的变换或替换,都应涵盖在本发明的包含范围之内,因此,本发明的保护范围应该以权利要求书的保护范围为准。 The above are merely specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art within the technical scope disclosed by the present invention, it is understood contemplated transformation or alternatively, both to be encompassed within the scope of the present invention comprises, therefore, the scope of the present invention should be protected by the scope of the claims and their equivalents. 结合这里披露的本发明的说明和实践,本发明的其他实施例对于本领域技术人员都是显而易见的。 Disclosed and described herein in conjunction with the practice of the invention, other embodiments of the present invention to those skilled in the art are apparent. 说明和实施例仅被认为是示例性的,本发明的真正范围和主旨均由权利要求所限定。 Specification and examples be considered as exemplary only, with a true scope and spirit of the invention as defined by the claims.

Claims (7)

1. 一种针对回光反射标志的可变权重像点定位方法,所述方法包括如下步骤: a) 设置多个物方标志点; b) 在不同站位对所述多个物方标志点进行拍摄,获取成像光斑上的多个位置点处的灰度值; c) 获取和背景噪声有关的第二权重因子0,其中所述第二权重因子0为所述成像光斑的平均灰度值,或所述第二权重因子e为〇; d) 获取与测量场环境变量相关的第一权重因子a,所述第一权重因子a的最优值的计算方法如下: d1)测量求出成像光斑上所述多个位置点的坐标,提取出求得的二维点中心坐标(xi(l,yi(l)作为真值; d2)使用包含下述公式的算法求出成像光斑上所述多个位置点坐标(Xi,yj)作为测量值,所述公式为: A variable weight for a return light reflected ghost point positioning marker, the method comprising the steps of: a) providing a plurality of landmark object side; b) a plurality of different stations of the object side to the mark point shooting to acquire gradation value at a plurality of points on the imaging position of the light spot; c) acquiring a second weight, and background noise associated weighting factor 0, wherein the second weighting factor 0 is the average gray value of the image spot or the second weighting factor e is square; D) obtaining a first weight associated with the environment variable measurement field a weighting factor, the weighting factor of a first method of calculating the optimum values ​​as follows: d1) measuring imaging obtained the coordinate points of the plurality of positions on the spot, the two-dimensional point extracting center coordinates obtained (xi (l, yi (l) as the true value; algorithm d2) containing the following equation was determined on the spot imaging a plurality of position coordinates (Xi, yj) as the measured value, the formula is:
Figure CN103411535BC00021
d3)求出所述测量值与所述真值的残差的均值; d4)取最小均值对应的a为最优值; e) 根据所获取的第一权重因子a和第二权重因子0,基于下述公式计算所测量椭圆形像点的点中心坐标(Xtl,yQ), d3) obtaining a residual mean value and the measured value of the true; D4) takes a minimum mean corresponding to an optimum value; E) according to the acquired first weight and a second weight factors weighting factor 0, is calculated based on the following formula ellipse point image center coordinates of the measured points (Xtl, yQ),
Figure CN103411535BC00022
其中xJPly』为成像光斑上的位置点的坐标(xi,yp,g(Xi,yp为(Xi,yj处成像光斑的灰度值,[W1^2]为成像光斑的宽度范围,[Vh2]为成像光斑的高度范围。 Wherein xJPly "coordinate position of a point on the imaging spot (xi, yp, g (Xi, yp is (Xi, yj gradation value of the image spot, [W1 ^ 2] is the width of the imaging spot, [Vh2] height range of the imaging spot.
2. 如权利要求1所述的像点定位方法,其中所述物方标志点为利用定向反光材料制作的圆形RRT标志。 2. The image point positioning method according to claim 1, wherein the object side, a circular mark point for the use of directional reflecting materials produced RRT flag.
3. 如权利要求2所述的像点定位方法,其中所述RRT标志的直径范围在3-12mm。 3. The method of locating the image point as claimed in claim 2, wherein a diameter ranging flag in the RRT 3-12mm.
4. 如权利要求3所述的像点定位方法,其中所述RRT标志的直径为6mm。 4. The method of locating the image point as claimed in claim 3, wherein the diameter of the RRT flag to 6mm.
5. 如权利要求1所述的像点定位方法,其中所述物方标志点均匀地分布在被测场中。 5. The method of locating the image point of claim 1, wherein the object side marker points uniformly distributed in the measured field.
6. 如权利要求1所述的像点定位方法,其中[W1,W2]和Di1,h2]取值范围分别略大于成像光斑的宽度Wtl和高度Iitl。 6. The method of locating the image point of claim 1, wherein [W1, W2], and Di1, h2] are in the range slightly larger than the width and height of the spot image Wtl Iitl.
7. 如权利要求1所述的像点定位方法,其中所述第一权重因子a的取值为整数。 7. The method of locating the image point of claim 1, wherein said first weight factor is a value of an integer.
CN201310340841.2A 2013-08-07 2013-08-07 A variable weight for a return light reflected ghost marker point positioning CN103411535B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310340841.2A CN103411535B (en) 2013-08-07 2013-08-07 A variable weight for a return light reflected ghost marker point positioning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310340841.2A CN103411535B (en) 2013-08-07 2013-08-07 A variable weight for a return light reflected ghost marker point positioning

Publications (2)

Publication Number Publication Date
CN103411535A CN103411535A (en) 2013-11-27
CN103411535B true CN103411535B (en) 2015-08-05

Family

ID=49604564

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310340841.2A CN103411535B (en) 2013-08-07 2013-08-07 A variable weight for a return light reflected ghost marker point positioning

Country Status (1)

Country Link
CN (1) CN103411535B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107525656B (en) * 2017-09-15 2019-07-26 上海汽车集团股份有限公司 Imaging independent positioning method of the illuminator in glass for vehicle window

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6076465A (en) * 1996-09-20 2000-06-20 Kla-Tencor Corporation System and method for determining reticle defect printability
CN101126638A (en) * 2007-09-29 2008-02-20 张小虎 Pick-up measuring method for checking road surface planeness
CN102003945A (en) * 2010-10-28 2011-04-06 汪远银 Virtual optical extensometer and measurement method thereof
CN102003938A (en) * 2010-10-11 2011-04-06 中国人民解放军信息工程大学 Thermal state on-site detection method for large high-temperature forging
CN102073324A (en) * 2010-12-29 2011-05-25 哈尔滨工业大学 Linearly polarized light-based polarization tracking system and method
CN102081738A (en) * 2011-01-06 2011-06-01 西北工业大学 Method for positioning mass center of spatial object star image
CN102840825A (en) * 2012-08-21 2012-12-26 华北电力大学 Particle locating system and method
CN203100681U (en) * 2013-03-22 2013-07-31 胡长华 Portable non-contact dimension measuring apparatus

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8068673B2 (en) * 2006-12-01 2011-11-29 Beihang University Rapid and high precision centroiding method and system for spots image
CN102696041B (en) * 2009-12-02 2016-03-02 塔塔咨询服务有限公司 Cost-effective for eye tracking and confirmation of driver drowsiness and robust system and method

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6076465A (en) * 1996-09-20 2000-06-20 Kla-Tencor Corporation System and method for determining reticle defect printability
CN101126638A (en) * 2007-09-29 2008-02-20 张小虎 Pick-up measuring method for checking road surface planeness
CN102003938A (en) * 2010-10-11 2011-04-06 中国人民解放军信息工程大学 Thermal state on-site detection method for large high-temperature forging
CN102003945A (en) * 2010-10-28 2011-04-06 汪远银 Virtual optical extensometer and measurement method thereof
CN102073324A (en) * 2010-12-29 2011-05-25 哈尔滨工业大学 Linearly polarized light-based polarization tracking system and method
CN102081738A (en) * 2011-01-06 2011-06-01 西北工业大学 Method for positioning mass center of spatial object star image
CN102840825A (en) * 2012-08-21 2012-12-26 华北电力大学 Particle locating system and method
CN203100681U (en) * 2013-03-22 2013-07-31 胡长华 Portable non-contact dimension measuring apparatus

Also Published As

Publication number Publication date
CN103411535A (en) 2013-11-27

Similar Documents

Publication Publication Date Title
Subbarao et al. Accurate recovery of three-dimensional shape from image focus
CN101825431B (en) Reference image techniques for three-dimensional sensing
KR20080111474A (en) Three-dimensional sensing using speckle patterns
US7324663B2 (en) Flight parameter measurement system
US6985183B2 (en) Method for exploring viewpoint and focal length of camera
US9410827B2 (en) Measurement using a calibration pattern
US9672623B2 (en) Image calibration
Pan et al. Optimization of a three-dimensional digital image correlation system for deformation measurements in extreme environments
EP1509781B1 (en) Flight parameter measurement system
CN101566461B (en) Method for quickly measuring blade of large-sized water turbine
CN101231750A (en) Calibrating method of binocular three-dimensional measuring system
JP2010259605A (en) Visual line measuring device and visual line measuring program
CN103292710A (en) Distance measuring method applying binocular visual parallax error distance-measuring principle
US6741279B1 (en) System and method for capturing document orientation information with a digital camera
Song et al. Use of LCD panel for calibrating structured-light-based range sensing system
Douxchamps et al. High-accuracy and robust localization of large control markers for geometric camera calibration
US9989952B2 (en) Image calibration
CN101577002B (en) Applied to the target detection fisheye imaging system calibration method
KR20130032368A (en) Three-dimensional measurement apparatus, three-dimensional measurement method, and storage medium
US20090033947A1 (en) Method for repeatable optical determination of object geometry dimensions and deviations
CN101308012B (en) Double monocular white light three-dimensional measuring systems calibration method
CN102183214B (en) Method for optically detecting large-aperture aspherical mirror structure
US20190186901A1 (en) Methods and Apparatus for Enhancing Depth Maps with Polarization Cues
JP5961945B2 (en) Image processing apparatus, projector and projector system having the image processing apparatus, image processing method, program thereof, and recording medium recording the program
JP4109075B2 (en) Method for measuring the rotation characteristics and flight characteristics of a sphere, and a device for measuring the rotation characteristics and flight characteristics of a sphere

Legal Events

Date Code Title Description
C06 Publication
C10 Request of examination as to substance
C14 Granted