CN112001379A - Correction algorithm of automobile instrument fixed viewpoint reading instrument based on machine vision - Google Patents
Correction algorithm of automobile instrument fixed viewpoint reading instrument based on machine vision Download PDFInfo
- Publication number
- CN112001379A CN112001379A CN202010425371.XA CN202010425371A CN112001379A CN 112001379 A CN112001379 A CN 112001379A CN 202010425371 A CN202010425371 A CN 202010425371A CN 112001379 A CN112001379 A CN 112001379A
- Authority
- CN
- China
- Prior art keywords
- point
- image
- instrument
- pointer
- coordinate system
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000012937 correction Methods 0.000 title claims abstract description 28
- 238000003384 imaging method Methods 0.000 claims abstract description 31
- 238000000034 method Methods 0.000 claims description 37
- 239000011159 matrix material Substances 0.000 claims description 27
- 239000002245 particle Substances 0.000 claims description 12
- 238000013519 translation Methods 0.000 claims description 12
- 230000003287 optical effect Effects 0.000 claims description 9
- 230000000694 effects Effects 0.000 claims description 3
- 238000004164 analytical calibration Methods 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 description 4
- 208000003464 asthenopia Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D18/00—Testing or calibrating apparatus or arrangements provided for in groups G01D1/00 - G01D15/00
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/80—Geometric correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/02—Recognising information on displays, dials, clocks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
本发明公开了一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法,具体包括如下步骤:建立仪表成像几何模型;求出所需的焦距f和物距H;将仪表图像进行二值化分割,得到分度线和指针;对指针和分度线进行标记,计算指针尖端Pu点的坐标(xu,yu),并根据Pu点的坐标,求O1Pu;根据指针和仪表盘的距离D,求长度PuPu1;在O1Pu连线上找出Pu1点,该点即为修正后指针的尖端Pu1(xu1,yu1),求出直线方程
并得出针旋转的角度;通过将指针旋转角度和最近分度线角度进行对比,可判别仪表示数。降低了对仪表校验控制系统的复杂性,本发明减少了由于视角偏移产生的读数误差。The invention discloses a machine vision-based correction algorithm for reading an instrument from a fixed viewpoint of an automobile instrument, which specifically includes the following steps: establishing an imaging geometric model of the instrument; finding the required focal length f and object distance H; Divide the division to get the index line and pointer; mark the pointer and the index line, calculate the coordinates (x u , y u ) of the point P u at the tip of the pointer, and find O 1 P u according to the coordinates of the point P u ; The distance D between the pointer and the instrument panel, find the length P u P u1 ; find the point P u1 on the O 1 P u connection line, which is the tip of the corrected pointer P u1 (x u1 , y u1 ), find out straight line equation
And the angle of needle rotation is obtained; by comparing the rotation angle of the pointer with the angle of the nearest graduation line, the number indicated by the instrument can be discriminated. The complexity of the instrument calibration control system is reduced, and the invention reduces the reading error caused by the deviation of the viewing angle.Description
技术领域technical field
本发明属于机器视觉检测技术领域,涉及一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法。The invention belongs to the technical field of machine vision detection, and relates to a machine vision-based correction algorithm for reading an instrument from a fixed viewpoint of an automobile instrument.
背景技术Background technique
汽车仪表显示汽车的各种数据,反馈汽车工作状态,对汽车的安全行驶起着重要作用。目前仪表检测采用传统的人工观测方式,由工作人员目测各仪表指针与刻度间的压线情况及各指示灯显示状态等来判断产品质量是否合格。人工检测受到人为观测角度、观测距离及人眼疲劳程度等主观因素影响,存在精度低,可靠性差,重复性差,检测时间长,效率低等一系列问题。因此,找到一种读取仪表示值不存在角度视觉误差、满足读表准则的方法是急需的。The car instrument displays various data of the car and feeds back the working status of the car, which plays an important role in the safe driving of the car. At present, the instrument detection adopts the traditional manual observation method, and the staff visually observes the pressure line between the pointer and the scale of each instrument and the display status of each indicator light to judge whether the product quality is qualified. Manual detection is affected by subjective factors such as artificial observation angle, observation distance, and human eye fatigue, and has a series of problems such as low accuracy, poor reliability, poor repeatability, long detection time, and low efficiency. Therefore, it is urgent to find a method that the reading instrument indicates that the value has no angular visual error and meets the reading criteria.
仪表校验是一项精密的测试工作,不论是数字仪表还是指针式仪表,均可以利用计算机视觉实现全自动校验。仪表自动校验系统要求当对每次输入量变化引起的指针偏转,示值判读时都需要移动摄像机视点,并要重新标定坐标系。这很大程度上增加了控制系统的复杂性、增加了校验时间,并且机械装置长时间使用可能会出现机械故障。Meter calibration is a precise test work. Whether it is a digital meter or an analog meter, computer vision can be used to achieve automatic calibration. The instrument automatic calibration system requires that the camera viewpoint needs to be moved and the coordinate system needs to be re-calibrated when the pointer is deflected and the indication value is interpreted each time the input quantity changes. This greatly increases the complexity of the control system, increases the calibration time, and mechanical failures may occur when the mechanical device is used for a long time.
发明内容SUMMARY OF THE INVENTION
本发明的目的是提供一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法,该算法降低了对仪表校验控制系统的复杂性,减少了由于视角偏移产生的读数误差。The purpose of the present invention is to provide a machine vision-based correction algorithm for reading the instrument from a fixed viewpoint of an automobile instrument, which reduces the complexity of the instrument calibration control system and reduces the reading error caused by the deviation of the viewing angle.
本发明所采用的技术方案是,一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法,具体包括如下步骤:The technical solution adopted in the present invention is a machine vision-based correction algorithm for reading an instrument from a fixed viewpoint of an automobile instrument, which specifically includes the following steps:
步骤1,利用摄像机的针孔模型表达世界坐标系、摄像机坐标系和图像坐标系之间的定量关系,并建立仪表成像几何模型;Step 1, use the pinhole model of the camera to express the quantitative relationship between the world coordinate system, the camera coordinate system and the image coordinate system, and establish an instrument imaging geometric model;
步骤2,利用径向平衡条件,对摄像机的标定采用Tsa i提出的二步法,求出所需的焦距f和物距H;Step 2, using the radial balance condition, adopt the two-step method proposed by Tsai for the calibration of the camera, and obtain the required focal length f and object distance H;
步骤3,将仪表图像进行二值化分割,得到分度线和指针;Step 3, the instrument image is binarized and segmented to obtain a graduation line and a pointer;
步骤4,对步骤3所得的指针和分度线进行标记,计算指针尖端Pu点的坐标(xu,yu),并根据Pu点的坐标,求O1Pu;Step 4, mark the pointer and graduation line obtained in step 3, calculate the coordinates (x u , y u ) of the point P u at the tip of the pointer, and find O 1 P u according to the coordinates of the point P u ;
步骤5,根据指针和仪表盘的距离D,求长度PuPu1;Step 5, according to the distance D between the pointer and the instrument panel, find the length P u P u1 ;
步骤6,根据步骤5所得的PuPu1,在O1Pu连线上找出Pu1点,该点即为修正后指针的尖端Pu1(xu1,yu1),求出直线方程并得出针旋转的角度;Step 6: According to the P u P u1 obtained in step 5, find the point P u1 on the O 1 P u connection line, which is the tip of the pointer after the correction P u1 (x u1 , y u1 ), and obtain the equation of the straight line And get the angle of needle rotation;
步骤7,通过将步骤6所得的指针旋转角度和最近分度线角度进行对比,可判别仪表示数。Step 7: By comparing the rotation angle of the pointer obtained in step 6 with the angle of the nearest graduation line, the number indicated by the instrument can be discriminated.
本发明的特点还在于,The present invention is also characterized in that,
步骤1的具体过程为:The specific process of step 1 is:
步骤1.1,由摄像机针孔成像模型,通过引入旋转矩阵R和平移矩阵T,建立参考坐标系和摄像机坐标系之间的关系:Step 1.1, from the camera pinhole imaging model, by introducing the rotation matrix R and the translation matrix T, the relationship between the reference coordinate system and the camera coordinate system is established:
式中,R为3X3的旋转矩阵,T为平移矩阵,(x,y,z)为摄像机坐标系下的坐标,(xw,yw,zw)为参考坐标系下的坐标;In the formula, R is the rotation matrix of 3X3, T is the translation matrix, (x, y, z) is the coordinate under the camera coordinate system, (x w , y w , z w ) is the coordinate under the reference coordinate system;
步骤1.2,建立图像坐标系和摄像机坐标系之间的关系,具体如下:Step 1.2, establish the relationship between the image coordinate system and the camera coordinate system, as follows:
式中,(X,Y,Z)为图像坐标系中的坐标,(X0,Y0)为图像坐标系中的任意一点坐标,dx为每一个像素在x轴上的尺寸,dy为每一个像素在y轴上的尺寸;In the formula, (X, Y, Z) are the coordinates in the image coordinate system, (X 0 , Y 0 ) are the coordinates of any point in the image coordinate system, d x is the size of each pixel on the x-axis, and dy is the size of each pixel on the y-axis;
步骤1.3,设表盘上的一点P在参考坐标系中为P(Xw,Yw,Zw),在摄像机坐标系坐标为P(x,y,z),该P点经摄像机成像后与像平面上理想像点Pu(xu,yu,f)对应,求理想投影点Pu(xu,yu,f)和畸变后的实际成像平面上的点Pd(xd,yd,f)之间的关系;Step 1.3, set a point P on the dial to be P(X w , Y w , Z w ) in the reference coordinate system and P(x, y, z) in the camera coordinate system. The ideal image point P u (x u , y u , f) on the image plane corresponds to the ideal projection point P u (x u , y u , f) and the distorted point on the actual imaging plane P d (x d , The relationship between y d ,f);
步骤1.4,根据仪表实际成像平面上的上点Pd(xd,yd)与计算机存储器中点S(uf,vf)之间的关系,求仪表成型几何模型。Step 1.4, according to the relationship between the upper point P d (x d , y d ) on the actual imaging plane of the instrument and the point S (u f , v f ) in the computer memory, obtain the geometric model of the instrument forming.
步骤1.3的具体过程如下:The specific process of step 1.3 is as follows:
步骤1.3.1,根据Pu(xu,yu,f)建立仪表成像方程:Step 1.3.1, establish the instrument imaging equation according to P u (x u , y u , f):
步骤1.3.2,确定图像焦距f和物距H之间的关系;Step 1.3.2, determine the relationship between the image focal length f and the object distance H;
Z=H+f (5);Z=H+f(5);
其中,Z为仪表平面在光轴方向上的坐标;Among them, Z is the coordinate of the instrument plane in the direction of the optical axis;
步骤1.3.3,考虑畸变对Pu点的影响,失真图像上的控制点为Pd(xd,yd,f),经图像采集输入到计算机存储器中,设Pd对应帧存图像中的S(uf,vf,f)点,建立矫正矩阵,具体如下:Step 1.3.3, consider the effect of distortion on the point P u , the control point on the distorted image is P d (x d , y d , f), which is input into the computer memory through image acquisition, and set P d corresponding to the frame memory image The S(u f ,v f ,f) points of , establish a correction matrix, as follows:
步骤1.3.4,由于畸变造成成像点位置的径向移动,使得理想投影点Pu(xu,yu,f)和畸变后的Pd(xd,yd,f)的关系为:Step 1.3.4, due to the radial movement of the imaging point position due to the distortion, the relationship between the ideal projection point P u (x u , y u , f) and the distorted P d (x d , y d , f) is:
其中,k1是畸变系数,且r=xd 2+yd 2。where k 1 is the distortion coefficient, and r=x d 2 +y d 2 .
步骤1.4的具体过程如下:The specific process of step 1.4 is as follows:
以像元数为单位,并引入线度单位到像素单位的比例因子δu和δv,分别表示出摄像机x方向和y方向两个相邻像元中心之间的距离,设(u0,v0)表示成像平面中心对应的计算机图像坐标,则Taking the number of pixels as the unit, and introducing the scaling factors δ u and δ v from the linear unit to the pixel unit, it represents the distance between the centers of two adjacent pixels in the x-direction and y-direction of the camera respectively. Let (u 0 , v 0 ) represents the computer image coordinates corresponding to the center of the imaging plane, then
结合式(1)至式(8),得到可表达P点坐标(Xw,Yw,Zw)与计算机存储器中图像S(uf,vf,f)之间对应关系Combining equations (1) to (8), the corresponding relationship between the coordinates (X w , Y w , Z w ) of the expressible point P and the image S (u f , v f , f) in the computer memory is obtained
其中in
r2=δu 2(uf-u0)2+δv 2(vf-u0)2 (10);r 2 =δ u 2 (u f -u 0 ) 2 +δ v 2 (v f -u 0 ) 2 (10);
r2即为最终建立的仪表成型几何模型。r 2 is the final instrument forming geometric model.
步骤2的具体过程如下:The specific process of step 2 is as follows:
步骤2.1,求P(x,y,z)点在坐标系中的坐标及摄像机的旋转矩阵R和平移矩阵T的值;Step 2.1, find the coordinates of the point P(x, y, z) in the coordinate system and the values of the camera's rotation matrix R and translation matrix T;
具体为:利用模板的黑圆点的空间坐标及径向平衡条件建立如下方程:Specifically, the following equations are established by using the space coordinates of the black dots of the template and the radial balance conditions:
结合公式(1)将x、y展开,得如下公式(12)、(13):Combining formula (1) to expand x and y, the following formulas (12) and (13) are obtained:
令Zw=0,代入实际N个点(xd,yd)的数据,求解方程组(12)、(13),即可得出摄像机的旋转矩阵R和平移矩阵T的值及P(x,y,z)点坐标值;Let Z w = 0, substitute the data of the actual N points (x d , y d ), and solve the equations (12) and (13), the values of the camera’s rotation matrix R and translation matrix T and P( x, y, z) point coordinate value;
步骤2.2,将步骤2.1所得结果代入公式(3)、(4)、(5)、(7)中,联立求解,即得畸变系数k1、焦距f和物距H。Step 2.2: Substitute the results obtained in step 2.1 into formulas (3), (4), (5), and (7), and solve simultaneously to obtain the distortion coefficient k 1 , the focal length f and the object distance H.
步骤3的具体过程为:The specific process of step 3 is:
步骤3.1,给定一个初始阈值Th=Th0,从头开始搜索,则将仪表图像原图分为C1和C2两类;Step 3.1, given an initial threshold Th =T h0 , start the search from the beginning, then divide the original image of the instrument image into two categories: C1 and C2;
步骤3.2,分别计算C1和C2两类图像的类内方差和均值;Step 3.2, calculate the intra-class variance and mean of C1 and C2 images respectively;
式中,f(x,y)为采集的图像;Nc1为像素被分在C1的概率;Nc2为像素被分在C2的概率;μ1为C1类图像的均值;μ2为C2类图像的均值;σ2 1为C1类图像的方差;σ2 2为C2类图像的方差;In the formula, f(x, y) is the collected image; N c1 is the probability that the pixel is classified in C1; N c2 is the probability that the pixel is classified in C2; μ 1 is the average value of the C1 class image; μ 2 is the C2 class The mean value of the image; σ 2 1 is the variance of the C1 class image; σ 2 2 is the variance of the C2 class image;
步骤3.3,对图像进行分类处理:如果|f(x,y)-μ1|≤|f(x,y)-μ2|,则f(x,y)属于C1,否则f(x,y)属于C2;Step 3.3, classify the image: if |f(x,y)-μ 1 |≤|f(x,y)-μ 2 |, then f(x, y) belongs to C1, otherwise f(x, y ) belongs to C2;
步骤3.4,对步骤3.3重新分类后得到的C1和C2中的像素,分别按照公式(14)~(17)重新计算各自的均值与方差;Step 3.4, for the pixels in C1 and C2 obtained after reclassification in step 3.3, recalculate their respective mean values and variances according to formulas (14) to (17);
步骤3.5,如果当前像素点的方差值满足如下关系:Step 3.5, if the variance value of the current pixel satisfies the following relationship:
则输出计算得到的阈值Th(t-1),否则重新选取像素点,重复执行步骤3.4~步骤3.5;Then output the calculated threshold Th (t-1), otherwise select the pixel again, and repeat steps 3.4 to 3.5;
步骤3.6,根据步骤3.5输出的阈值对图像进行分类,得到只有分度线和指针的黑白图像。Step 3.6, classify the image according to the threshold output in step 3.5, and obtain a black and white image with only graduation lines and pointers.
步骤4的具体过程为:The specific process of step 4 is:
假设二值图像中为0的点是背景,为1的点是微粒,早算法中采用八邻接点搜寻,算法如下:Assuming that the point of 0 in the binary image is the background, and the point of 1 is the particle, the early algorithm uses eight adjacent points to search, and the algorithm is as follows:
(1)算法开始,令标记Label=1;(1) At the beginning of the algorithm, let the mark Label=1;
(2)自左向右、自上而下扫描图像,找寻其值为1的种子点,设定种子点标记=Label;假如找不到种子点,则结束整个标记算法;(2) scan the image from left to right, from top to bottom, find the seed point whose value is 1, set the seed point mark=Label; if the seed point cannot be found, then end the entire labeling algorithm;
(3)对于种子点周围同值之像素点做下述操作:(3) Do the following for the pixels of the same value around the seed point:
x轴方向x-axis direction
(a)自左向右逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1的像素点标记=Label;(a) Scan the image point by point from left to right, if f(x, y) is marked as Label, then let f(x, y) eight adjacent points in the pixel point with a value of 1 mark = Label;
(b)自右向左逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1像素点标记=Label;(b) Scan the image point by point from right to left, if f(x, y) is marked as Label, then let f(x, y) in the eight adjacent points be 1 pixel point mark=Label;
y轴方向y-axis direction
(c)由上往下逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1的像素点标记=Label;(c) Scan the image point by point from top to bottom, if f(x,y) is marked as Label, then let f(x,y) eight adjacent points in the pixel point with a value of 1 mark=Label;
(d)自下向上逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1的像素点标记=Label;(d) Scan the image point by point from bottom to top, if f(x, y) is marked as Label, then let f(x, y) eight adjacent points in the pixel point with a value of 1 mark = Label;
(4)经过四个方向扫描后,标记为L之微粒被完整取出,指定新的标记Label++重复步骤(2),直到标记完所有微粒;(4) After scanning in four directions, the particles marked L are completely taken out, and a new label Label++ is designated to repeat step (2) until all particles are marked;
(5)输出距离最远的微粒坐标作为Pu点的坐标;(5) Output the coordinates of the particles with the farthest distance as the coordinates of the P u point;
根据Pu点的坐标即可求得O1Pu为:According to the coordinates of the P u point, O 1 P u can be obtained as:
步骤5的具体过程如下:The specific process of step 5 is as follows:
假设P1是P2在标度盘平面上的垂直投影,在像平面上对应Pu1点,但读表准则要求P1点与P2点在像平面上应为同一点;Suppose that P 1 is the vertical projection of P 2 on the plane of the scale, and corresponds to the point P u1 on the image plane, but the meter reading criterion requires that point P 1 and point P 2 should be the same point on the image plane;
由几何关系可得三角形OcPP1和OPuPu1,以及OcPuO1和P2PP1是相似三角形,可找到图像的Pu1点,并求出PuPu1的距离。From the geometric relationship, triangles O c PP 1 and OP u P u1 can be obtained, and O c P u O 1 and P 2 PP 1 are similar triangles. The P u1 point of the image can be found and the distance of P u P u1 can be obtained.
具体实现过程如下:The specific implementation process is as follows:
步骤6的具体过程如下:The specific process of step 6 is as follows:
步骤6.1,在O1Pu连线上找出距离Pu点长度为PuPu1的点,此即为修正后指针的尖端Pu1(xu1,yu1);Step 6.1, on the O 1 P u connection line, find a point with a length of P u P u1 from the point P u , which is the tip of the corrected pointer P u1 (x u1 , y u1 );
步骤6.2,根据步骤6.1所得结果求出直线方程为并得出指针旋转的角度 Step 6.2, according to the result obtained in step 6.1, the equation of the straight line is obtained as and get the angle of pointer rotation
步骤7的具体过程如下:The specific process of step 7 is as follows:
步骤7.1,将修正后的仪表指针位置绘在原图像背景上,通过在视点固定时的重构图像,形成一幅用于精确读数的新图像;Step 7.1, draw the corrected instrument pointer position on the background of the original image, and form a new image for accurate reading through the reconstructed image when the viewpoint is fixed;
步骤7.2,将步骤6所得的指针旋转角度与最近分度线角度进行对比,判别仪表示数。Step 7.2, compare the rotation angle of the pointer obtained in step 6 with the angle of the nearest graduation line, and the discriminator indicates the number.
本发明的有益效果是:本发明提供一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法,该算法利用径向平衡条件,采用二步法进行摄像机的标定,通过修正的标记算法来确定指针尖端,进而重构图像与原图像形成对比,解决仪表示数准确读取的问题。该方法降低了对仪表校验控制系统的复杂性,减少了由于视角偏移产生的读数误差,并在仪表示数读取中具有广泛的应用。The beneficial effects of the present invention are as follows: the present invention provides a machine vision-based correction algorithm for reading instruments from a fixed viewpoint of an automobile instrument. The algorithm utilizes the radial balance condition, adopts a two-step method to calibrate the camera, and uses the modified marking algorithm to calibrate the camera. Determine the tip of the pointer, and then the reconstructed image is compared with the original image to solve the problem of accurate reading of the meter. The method reduces the complexity of the meter calibration control system, reduces the reading error caused by the deviation of the viewing angle, and has a wide range of applications in the meter reading.
附图说明Description of drawings
图1是本发明一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法中建立的仪表成像几何模型图;Fig. 1 is a kind of instrument imaging geometric model diagram established in a machine vision-based vehicle instrument fixed viewpoint reading instrument correction algorithm of the present invention;
图2是本发明一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法中建立的参数校正网络模板;Fig. 2 is a parameter correction network template established in a machine vision-based vehicle meter fixed viewpoint reading meter correction algorithm of the present invention;
图3是本发明一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法中仪表示数读取的几何修正模型。3 is a geometric correction model of meter reading in a machine vision-based correction algorithm for reading meters from a fixed viewpoint of an automobile meter according to the present invention.
具体实施方式Detailed ways
下面结合附图和具体实施方式对本发明进行详细说明。The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
本发明一种基于机器视觉的汽车仪表固定视点读取仪表的修正算法,具体包括如下步骤:The present invention is a machine vision-based correction algorithm for reading instruments from a fixed viewpoint of an automobile instrument, which specifically includes the following steps:
步骤1,利用摄像机的针孔模型表达世界坐标系、摄像机坐标系和图像坐标系之间的定量关系,并建立仪表成像几何模型;Step 1, use the pinhole model of the camera to express the quantitative relationship between the world coordinate system, the camera coordinate system and the image coordinate system, and establish an instrument imaging geometric model;
步骤1的具体过程为:The specific process of step 1 is:
步骤1.1,由摄像机针孔成像模型,通过引入旋转矩阵R和平移矩阵T,建立参考坐标系和摄像机坐标系之间的关系:Step 1.1, from the camera pinhole imaging model, by introducing the rotation matrix R and the translation matrix T, the relationship between the reference coordinate system and the camera coordinate system is established:
式中,R为3X3的旋转矩阵,T为平移矩阵,(x,y,z)为摄像机坐标系下的坐标,(xw,yw,zw)为参考坐标系下的坐标;In the formula, R is the rotation matrix of 3X3, T is the translation matrix, (x, y, z) is the coordinate under the camera coordinate system, (x w , y w , z w ) is the coordinate under the reference coordinate system;
步骤1.2,建立图像坐标系和摄像机坐标系之间的关系,具体如下:Step 1.2, establish the relationship between the image coordinate system and the camera coordinate system, as follows:
式中,(X,Y,Z)为图像坐标系中的坐标,(X0,Y0)为图像坐标系中的任意一点坐标,dx为每一个像素在x轴上的尺寸,dy为每一个像素在y轴上的尺寸;In the formula, (X, Y, Z) are the coordinates in the image coordinate system, (X 0 , Y 0 ) are the coordinates of any point in the image coordinate system, d x is the size of each pixel on the x-axis, and dy is the size of each pixel on the y-axis;
使用世界坐标系、摄像机坐标系和图像坐标系之间的关系,建立如图1仪表成像几何模型。Using the relationship between the world coordinate system, the camera coordinate system and the image coordinate system, the imaging geometric model of the instrument as shown in Figure 1 is established.
以世界坐标系(Ow,Xw,Yw,Zw)作为参考坐标系,其原点Ow的坐标固定不变。由点Oc与Xc,Yc,Zc轴组成的直角坐标系为摄像机坐标系,点Oc为摄像机的光学中心,Xc—Yc平面与CCD成像平面平行。Zc轴即为光轴,它与CCD摄像机的图像平面垂直,其与图像平面的交点O1为图像平面坐标系原点,OcO1为摄像机的焦距f。取空间仪表表盘平面与图像平面平行,与光轴交点为O′,与图像平面距离即为物距H。Taking the world coordinate system (O w , X w , Y w , Z w ) as the reference coordinate system, the coordinates of its origin O w are fixed. The Cartesian coordinate system composed of point O c and X c , Y c , and Z c axes is the camera coordinate system, point O c is the optical center of the camera, and the X c -Y c plane is parallel to the CCD imaging plane. The Z c axis is the optical axis, which is perpendicular to the image plane of the CCD camera. The intersection O1 with the image plane is the origin of the image plane coordinate system, and OcO1 is the focal length f of the camera. Take the space instrument panel plane parallel to the image plane, the intersection point with the optical axis is O', and the distance from the image plane is the object distance H.
步骤1.3,设表盘上的一点P在参考坐标系中为P(Xw,Yw,Zw),在摄像机坐标系坐标为P(x,y,z),该P点经摄像机成像后与像平面上理想像点Pu(xu,yu,f)对应,求理想投影点Pu(xu,yu,f)和畸变后的实际成像平面上的点Pd(xd,yd,f)之间的关系;Step 1.3, set a point P on the dial to be P(X w , Y w , Z w ) in the reference coordinate system and P(x, y, z) in the camera coordinate system. The ideal image point P u (x u , y u , f) on the image plane corresponds to the ideal projection point P u (x u , y u , f) and the distorted point on the actual imaging plane P d (x d , The relationship between y d ,f);
根据摄像机、仪表变盘所处的位置,从上到下建立分别为世界坐标系、图像坐标系、摄像机坐标系三个坐标系。表盘上的一点P在参考坐标系中为P(Xw,Yw,Zw),在摄像机坐标系坐标为P(x,y,z),该点经摄像机成像后与像平面上理想像点Pu(xu,yu,f)对应。According to the position of the camera and the instrument panel, three coordinate systems are established from top to bottom: the world coordinate system, the image coordinate system, and the camera coordinate system. A point P on the dial is P(X w , Y w , Z w ) in the reference coordinate system and P(x, y, z) in the camera coordinate system. After imaging by the camera, this point is the ideal image on the image plane. The point P u (x u , y u , f) corresponds to.
步骤1.3的具体过程如下:The specific process of step 1.3 is as follows:
步骤1.3.1,根据Pu(xu,yu,f)建立仪表成像方程:Step 1.3.1, establish the instrument imaging equation according to P u (x u , y u , f):
步骤1.3.2,确定图像焦距f和物距H之间的关系;Step 1.3.2, determine the relationship between the image focal length f and the object distance H;
Z=H+f (5);Z=H+f(5);
其中,Z为仪表平面在光轴方向上的坐标;Among them, Z is the coordinate of the instrument plane in the direction of the optical axis;
步骤1.3.3,考虑畸变对Pu点的影响,失真图像上的控制点为Pd(xd,yd,f),经图像采集输入到计算机存储器中,设Pd对应帧存图像中的S(uf,vf,f)点,建立矫正矩阵,具体如下:Step 1.3.3, consider the effect of distortion on the point P u , the control point on the distorted image is P d (x d , y d , f), which is input into the computer memory through image acquisition, and set P d corresponding to the frame memory image The S(u f ,v f ,f) points of , establish a correction matrix, as follows:
步骤1.3.4,由于畸变造成成像点位置的径向移动,使得理想投影点Pu(xu,yu,f)和畸变后的Pd(xd,yd,f)的关系为:Step 1.3.4, due to the radial movement of the imaging point position due to the distortion, the relationship between the ideal projection point P u (x u , y u , f) and the distorted P d (x d , y d , f) is:
其中,k1是畸变系数,且r=xd 2+yd 2。where k 1 is the distortion coefficient, and r=x d 2 +y d 2 .
步骤1.4,根据仪表实际成像平面上的上点Pd(xd,yd)与计算机存储器中点S(uf,vf)之间的关系,求仪表成型几何模型。Step 1.4, according to the relationship between the upper point P d (x d , y d ) on the actual imaging plane of the instrument and the point S (u f , v f ) in the computer memory, obtain the geometric model of the instrument forming.
步骤1.4的具体过程如下:The specific process of step 1.4 is as follows:
以像元数为单位,并引入线度单位到像素单位的比例因子δu和δv,分别表示出摄像机x方向和y方向两个相邻像元中心之间的距离,设(u0,v0)表示成像平面中心对应的计算机图像坐标,则Taking the number of pixels as the unit, and introducing the scaling factors δ u and δ v from the linear unit to the pixel unit, it represents the distance between the centers of two adjacent pixels in the x-direction and y-direction of the camera respectively. Let (u 0 , v 0 ) represents the computer image coordinates corresponding to the center of the imaging plane, then
结合式(1)至式(8),得到可表达P点坐标(Xw,Yw,Zw)与计算机存储器中图像S(uf,vf,f)之间对应关系Combining equations (1) to (8), the corresponding relationship between the coordinates (X w , Y w , Z w ) of the expressible point P and the image S (u f , v f , f) in the computer memory is obtained
其中in
r2=δu 2(uf-u0)2+δv 2(vf-u0)2 (10);r 2 =δ u 2 (u f -u 0 ) 2 +δ v 2 (v f -u 0 ) 2 (10);
r2即为最终建立的仪表成型几何模型。r 2 is the final instrument forming geometric model.
步骤2,利用径向平衡条件,对摄像机的标定采用Tsa i提出的二步法,求出所需的焦距f和物距H;Step 2, using the radial balance condition, adopt the two-step method proposed by Tsai for the calibration of the camera, and obtain the required focal length f and object distance H;
根据CCD摄像机的参数,得到图像中心(u0,v0)对应的坐标,然后通过网络模板的先验知识及其成像数据,求出摄像机的内外参数,进而得出所需的焦距f和物距H。According to the parameters of the CCD camera, the coordinates corresponding to the image center (u 0 , v 0 ) are obtained, and then the internal and external parameters of the camera are obtained through the prior knowledge of the network template and its imaging data, and then the required focal length f and object are obtained. from H.
实际的参照模板选用白色背景下均匀分布的N个黑圆点,如图2所示,在像平面上投影后,经图像处理得到的N个点代表着成像的坐标关系。利用模板的黑圆点的空间坐标,加上径向平衡条件(直线O1Pd与O′P平行,对应矢量的叉积为零);The actual reference template selects N black dots evenly distributed under a white background, as shown in Figure 2, after being projected on the image plane, the N dots obtained through image processing represent the coordinate relationship of imaging. Using the space coordinates of the black dots of the template, plus the radial balance condition (the straight line O 1 P d is parallel to O'P, and the cross product of the corresponding vector is zero);
步骤2.1,求P(x,y,z)点在坐标系中的坐标及摄像机的旋转矩阵R和平移矩阵T的值;Step 2.1, find the coordinates of the point P(x, y, z) in the coordinate system and the values of the camera's rotation matrix R and translation matrix T;
建立如下方程:Create the following equation:
结合公式(1)将x、y展开,得如下公式(12)、(13):Combining formula (1) to expand x and y, the following formulas (12) and (13) are obtained:
令Zw=0,代入实际N个点(xd,yd)的数据,求解方程组(12)、(13),即可得出摄像机的旋转矩阵R和平移矩阵T的值及P(x,y,z)点坐标值;N≥6;Let Z w = 0, substitute the data of the actual N points (x d , y d ), and solve the equations (12) and (13), the values of the camera’s rotation matrix R and translation matrix T and P( x, y, z) point coordinate value; N≥6;
步骤2.2,将步骤2.1所得结果代入公式(3)、(4)、(5)、(7)中,联立求解,即得畸变系数k1、焦距f和物距H。Step 2.2: Substitute the results obtained in step 2.1 into formulas (3), (4), (5), and (7), and solve simultaneously to obtain the distortion coefficient k 1 , the focal length f and the object distance H.
步骤3,将仪表图像进行二值化分割,得到分度线和指针;Step 3, the instrument image is binarized and segmented to obtain a graduation line and a pointer;
步骤3的具体过程为:The specific process of step 3 is:
步骤3.1,给定一个初始阈值Th=Th0,从头开始搜索,则将仪表图像原图分为C1和C2两类;Step 3.1, given an initial threshold Th =T h0 , start the search from the beginning, then divide the original image of the instrument image into two categories: C1 and C2;
步骤3.2,分别计算C1和C2两类图像的类内方差和均值;Step 3.2, calculate the intra-class variance and mean of C1 and C2 images respectively;
式中,f(x,y)为采集的图像;Nc1为像素被分在C1的概率;Nc2为像素被分在C2的概率;μ1为C1类图像的均值;μ2为C2类图像的均值;σ2 1为C1类图像的方差;σ2 2为C2类图像的方差;In the formula, f(x, y) is the collected image; N c1 is the probability that the pixel is classified in C1; N c2 is the probability that the pixel is classified in C2; μ 1 is the average value of the C1 class image; μ 2 is the C2 class The mean value of the image; σ 2 1 is the variance of the C1 class image; σ 2 2 is the variance of the C2 class image;
式中,Nimage为像素在该图像的概率;p1为C1类像素在图像中的分布概率;p2为C2类像素在图像中的分布概率。In the formula, N image is the probability of pixels in the image; p 1 is the distribution probability of C1 class pixels in the image; p 2 is the distribution probability of C2 class pixels in the image.
步骤3.3,对图像进行分类处理:如果|f(x,y)-μ1|≤|f(x,y)-μ2|,则f(x,y)属于C1,否则f(x,y)属于C2;Step 3.3, classify the image: if |f(x,y)-μ 1 |≤|f(x,y)-μ 2 |, then f(x, y) belongs to C1, otherwise f(x, y ) belongs to C2;
步骤3.4,对步骤3.3重新分类后得到的C1和C2中的像素,分别按照公式(14)~(17)重新计算各自的均值与方差;Step 3.4, for the pixels in C1 and C2 obtained after reclassification in step 3.3, recalculate their respective mean values and variances according to formulas (14) to (17);
步骤3.5,如果当前像素点的方差值满足如下关系:Step 3.5, if the variance value of the current pixel satisfies the following relationship:
则输出计算得到的阈值Th(t-1),否则重新选取像素点,重复执行步骤3.4~步骤3.5;Then output the calculated threshold Th (t-1), otherwise select the pixel again, and repeat steps 3.4 to 3.5;
步骤3.6,根据步骤3.5输出的阈值对图像进行分类,得到只有分度线和指针的黑白图像。Step 3.6, classify the image according to the threshold output in step 3.5, and obtain a black and white image with only graduation lines and pointers.
步骤4,对步骤3所得的指针和分度线进行标记,计算指针尖端Pu点的坐标(xu,yu),并根据Pu点的坐标,求O1Pu;Step 4, mark the pointer and graduation line obtained in step 3, calculate the coordinates (x u , y u ) of the point P u at the tip of the pointer, and find O 1 P u according to the coordinates of the point P u ;
如图3所示,以世界坐标系(Ow,Xw,Yw,Zw)作为参考坐标系,其原点Ow的坐标固定不变。由点Oc与Xc,Yc,Zc轴组成的直角坐标系为摄像机坐标系,点Oc为摄像机的光学中心,Xc—Yc平面与CCD成像平面平行。Zc轴即为光轴,它与CCD摄像机的图像平面垂直,其与图像平面的交点O1为图像平面坐标系原点,OcO1为摄像机的焦距f。取空间仪表表盘平面与图像平面平行,与光轴交点为O′,与图像平面距离即为物距H。设仪表指针与标度盘距离设为D,O1XY是成像平面,P2为指针的尖端位置,Pu为其在像平面投影。As shown in FIG. 3 , the world coordinate system (O w , X w , Y w , Z w ) is used as the reference coordinate system, and the coordinates of its origin O w are fixed. The Cartesian coordinate system composed of point O c and X c , Y c , and Z c axes is the camera coordinate system, point O c is the optical center of the camera, and the X c -Y c plane is parallel to the CCD imaging plane. The Z c axis is the optical axis, which is perpendicular to the image plane of the CCD camera. The intersection O1 with the image plane is the origin of the image plane coordinate system, and OcO1 is the focal length f of the camera. Take the space instrument panel plane parallel to the image plane, the intersection point with the optical axis is O', and the distance from the image plane is the object distance H. Set the distance between the instrument pointer and the dial as D, O 1 XY is the imaging plane, P 2 is the position of the tip of the pointer, and Pu is its projection on the image plane.
步骤4的具体过程为:The specific process of step 4 is:
假设二值图像中为0的点是背景,为1的点是微粒,在算法中采用八邻接点搜寻,算法如下:Assuming that the point of 0 in the binary image is the background, and the point of 1 is the particle, the algorithm uses eight adjacent points to search, and the algorithm is as follows:
(1)算法开始,令标记Label=1;(1) At the beginning of the algorithm, let the mark Label=1;
(2)自左向右、自上而下扫描图像,找寻其值为1的种子点,设定种子点标记=Label。假如找不到种子点,则结束整个标记算法;(2) Scan the image from left to right and from top to bottom, find the seed point whose value is 1, and set the seed point mark=Label. If the seed point is not found, the whole marking algorithm ends;
(3)对于种子点周围同值之像素点做下述操作:(3) Do the following for the pixels of the same value around the seed point:
x轴方向x-axis direction
(a)自左向右逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1的像素点标记=Label;(a) Scan the image point by point from left to right, if f(x, y) is marked as Label, then let f(x, y) eight adjacent points in the pixel point with a value of 1 mark = Label;
(b)自右向左逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1像素点标记=Label;(b) Scan the image point by point from right to left, if f(x, y) is marked as Label, then let f(x, y) in the eight adjacent points be 1 pixel point mark=Label;
y轴方向y-axis direction
(c)由上往下逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1的像素点标记=Label;(c) Scan the image point by point from top to bottom, if f(x,y) is marked as Label, then let f(x,y) eight adjacent points in the pixel point with a value of 1 mark=Label;
(d)自下向上逐点扫描图像,若f(x,y)标记为Label,则令f(x,y)八邻接点中值为1的像素点标记=Label;(d) Scan the image point by point from bottom to top, if f(x, y) is marked as Label, then let f(x, y) eight adjacent points in the pixel point with a value of 1 mark = Label;
(4)经过四个方向扫描后,标记为L之微粒被完整取出,指定新的标记Label++重复步骤(2),直到标记完所有微粒;(4) After scanning in four directions, the particles marked L are completely taken out, and a new label Label++ is designated to repeat step (2) until all particles are marked;
(5)输出距离最远的微粒坐标作为Pu点的坐标;(5) Output the coordinates of the particles with the farthest distance as the coordinates of the P u point;
根据Pu点的坐标即可求得O1Pu为:According to the coordinates of the P u point, O 1 P u can be obtained as:
步骤5,根据指针和仪表盘的距离D,求长度PuPu1;Step 5, according to the distance D between the pointer and the instrument panel, find the length P u P u1 ;
步骤5的具体过程如下:The specific process of step 5 is as follows:
假设P1是P2在标度盘平面上的垂直投影,在像平面上对应Pu1点(成像时Pu1不存在),但读表准则要求P1点与P2点在像平面上应为同一点;Assuming that P 1 is the vertical projection of P 2 on the scale plane, it corresponds to the P u1 point on the image plane (P u1 does not exist during imaging), but the meter reading criterion requires that the P 1 point and the P 2 point should be on the image plane. for the same point;
由几何关系可得三角形OcPP1和OPuPu1,以及OcPuO1和P2PP1是相似三角形,可找到图像的Pu1点,并求出PuPu1的距离。From the geometric relationship, triangles O c PP 1 and OP u P u1 can be obtained, and O c P u O 1 and P 2 PP 1 are similar triangles. The P u1 point of the image can be found and the distance of P u P u1 can be obtained.
具体实现过程如下:The specific implementation process is as follows:
步骤6,根据步骤5所得的PuPu1,在O1Pu连线上找出Pu1点,该点即为修正后指针的尖端Pu1(xu1,yu1),求出直线方程并得出指针旋转的角度;Step 6: According to the P u P u1 obtained in step 5, find the point P u1 on the O 1 P u connection line, which is the tip of the pointer after the correction P u1 (x u1 , y u1 ), and obtain the equation of the straight line And get the angle of pointer rotation;
步骤6的具体过程如下:The specific process of step 6 is as follows:
步骤6.1,在O1Pu连线上找出距离Pu点长度为PuPu1的点,此即为修正后指针的尖端Pu1(xu1,yu1);Step 6.1, on the O 1 P u connection line, find a point with a length of P u P u1 from the point P u , which is the tip of the corrected pointer P u1 (x u1 , y u1 );
步骤6.2,根据步骤6.1所得结果求出直线方程为并得出指针旋转的角度 Step 6.2, according to the result obtained in step 6.1, the equation of the straight line is obtained as and get the angle of pointer rotation
步骤7,通过将步骤6所得的指针旋转角度和最近分度线角度进行对比,可判别仪表示数。Step 7: By comparing the rotation angle of the pointer obtained in step 6 with the angle of the nearest graduation line, the number indicated by the instrument can be discriminated.
步骤7的具体过程如下:The specific process of step 7 is as follows:
步骤7.1,将修正后的仪表指针位置绘在原图像背景上,通过在视点固定时的重构图像,形成一幅用于精确读数的新图像;Step 7.1, draw the corrected instrument pointer position on the background of the original image, and form a new image for accurate reading through the reconstructed image when the viewpoint is fixed;
步骤7.2,将步骤6所得的指针旋转角度与最近分度线角度进行对比,判别仪表示数。Step 7.2, compare the rotation angle of the pointer obtained in step 6 with the angle of the nearest graduation line, and the discriminator indicates the number.
Claims (10)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010425371.XA CN112001379A (en) | 2020-05-19 | 2020-05-19 | Correction algorithm of automobile instrument fixed viewpoint reading instrument based on machine vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010425371.XA CN112001379A (en) | 2020-05-19 | 2020-05-19 | Correction algorithm of automobile instrument fixed viewpoint reading instrument based on machine vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112001379A true CN112001379A (en) | 2020-11-27 |
Family
ID=73461484
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010425371.XA Pending CN112001379A (en) | 2020-05-19 | 2020-05-19 | Correction algorithm of automobile instrument fixed viewpoint reading instrument based on machine vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112001379A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112668578A (en) * | 2020-12-31 | 2021-04-16 | 中广核研究院有限公司 | Pointer instrument reading method and device, computer equipment and storage medium |
Citations (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08201021A (en) * | 1995-01-23 | 1996-08-09 | Mazda Motor Corp | Calibration method |
JP2005032028A (en) * | 2003-07-07 | 2005-02-03 | Ntt Power & Building Facilities Inc | Method for reading indicator value of indicator needle rotating meter, device for reading indicator value of indicator needle rotating meter, and program for reading meter indicator value |
CN102693543A (en) * | 2012-05-21 | 2012-09-26 | 南开大学 | Method for automatically calibrating Pan-Tilt-Zoom in outdoor environments |
CN102799867A (en) * | 2012-07-09 | 2012-11-28 | 哈尔滨工业大学 | Meter pointer angle identification method based on image processing |
CN103714329A (en) * | 2013-12-31 | 2014-04-09 | 长安大学 | Detecting algorithm for identifying meter needle |
CN105096317A (en) * | 2015-07-03 | 2015-11-25 | 吴晓军 | Fully automatic calibration method for high performance camera under complicated background |
CN105740829A (en) * | 2016-02-02 | 2016-07-06 | 暨南大学 | Scanning line processing based automatic reading method for pointer instrument |
CN105740856A (en) * | 2016-01-28 | 2016-07-06 | 宁波理工监测科技股份有限公司 | Method for reading readings of pointer instrument based on machine vision |
CN106570906A (en) * | 2016-11-09 | 2017-04-19 | 东南大学 | Rectangular pattern-based method for detecting distances under camera angle deflection condition |
CN106599897A (en) * | 2016-12-09 | 2017-04-26 | 广州供电局有限公司 | Machine vision-based pointer type meter reading recognition method and device |
CN107014312A (en) * | 2017-04-25 | 2017-08-04 | 西安交通大学 | A kind of integral calibrating method of mirror-vibrating line laser structured light three-dimension measuring system |
CN107085853A (en) * | 2017-05-04 | 2017-08-22 | 中国矿业大学 | Monitoring method of headframe deformation in mining area with monocular stereo vision of guide rail |
CN107167169A (en) * | 2017-07-03 | 2017-09-15 | 吉林大学 | Readings of pointer type meters identification measuring method based on NI Vision Builder for Automated Inspection |
CN107167172A (en) * | 2017-04-17 | 2017-09-15 | 江苏大学 | A kind of on-line monitoring method of bus type automobile digital instrument pointer functionality |
CN107463931A (en) * | 2017-07-06 | 2017-12-12 | 国家电网公司 | A kind of real-time pointer instrument reading method and device based on ARM platforms |
CN108009535A (en) * | 2017-11-21 | 2018-05-08 | 武汉中元华电科技股份有限公司 | A kind of simple pointer meter reading method based on machine vision |
CN109558871A (en) * | 2018-10-26 | 2019-04-02 | 中国科学院长春光学精密机械与物理研究所 | A kind of readings of pointer type meters recognition methods and device |
CN110189314A (en) * | 2019-05-28 | 2019-08-30 | 长春大学 | Image positioning method of automobile instrument panel based on machine vision |
CN110414510A (en) * | 2019-07-26 | 2019-11-05 | 华中科技大学 | A method for correcting readings of pointer instruments |
US20200034987A1 (en) * | 2018-07-25 | 2020-01-30 | Beijing Smarter Eye Technology Co. Ltd. | Method and device for building camera imaging model, and automated driving system for vehicle |
CN110852213A (en) * | 2019-10-30 | 2020-02-28 | 天津大学 | Multi-condition automatic reading method of pointer meter based on template matching |
CN111080711A (en) * | 2019-12-05 | 2020-04-28 | 东南大学 | A Magnification-Based Calibration Method of Microscopic Imaging System in Nearly Parallel State |
CN111091076A (en) * | 2019-12-03 | 2020-05-01 | 西北工业大学 | Measurement method of tunnel boundary data based on stereo vision |
-
2020
- 2020-05-19 CN CN202010425371.XA patent/CN112001379A/en active Pending
Patent Citations (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08201021A (en) * | 1995-01-23 | 1996-08-09 | Mazda Motor Corp | Calibration method |
JP2005032028A (en) * | 2003-07-07 | 2005-02-03 | Ntt Power & Building Facilities Inc | Method for reading indicator value of indicator needle rotating meter, device for reading indicator value of indicator needle rotating meter, and program for reading meter indicator value |
CN102693543A (en) * | 2012-05-21 | 2012-09-26 | 南开大学 | Method for automatically calibrating Pan-Tilt-Zoom in outdoor environments |
CN102799867A (en) * | 2012-07-09 | 2012-11-28 | 哈尔滨工业大学 | Meter pointer angle identification method based on image processing |
CN103714329A (en) * | 2013-12-31 | 2014-04-09 | 长安大学 | Detecting algorithm for identifying meter needle |
CN105096317A (en) * | 2015-07-03 | 2015-11-25 | 吴晓军 | Fully automatic calibration method for high performance camera under complicated background |
CN105740856A (en) * | 2016-01-28 | 2016-07-06 | 宁波理工监测科技股份有限公司 | Method for reading readings of pointer instrument based on machine vision |
CN105740829A (en) * | 2016-02-02 | 2016-07-06 | 暨南大学 | Scanning line processing based automatic reading method for pointer instrument |
CN106570906A (en) * | 2016-11-09 | 2017-04-19 | 东南大学 | Rectangular pattern-based method for detecting distances under camera angle deflection condition |
CN106599897A (en) * | 2016-12-09 | 2017-04-26 | 广州供电局有限公司 | Machine vision-based pointer type meter reading recognition method and device |
CN107167172A (en) * | 2017-04-17 | 2017-09-15 | 江苏大学 | A kind of on-line monitoring method of bus type automobile digital instrument pointer functionality |
CN107014312A (en) * | 2017-04-25 | 2017-08-04 | 西安交通大学 | A kind of integral calibrating method of mirror-vibrating line laser structured light three-dimension measuring system |
CN107085853A (en) * | 2017-05-04 | 2017-08-22 | 中国矿业大学 | Monitoring method of headframe deformation in mining area with monocular stereo vision of guide rail |
CN107167169A (en) * | 2017-07-03 | 2017-09-15 | 吉林大学 | Readings of pointer type meters identification measuring method based on NI Vision Builder for Automated Inspection |
CN107463931A (en) * | 2017-07-06 | 2017-12-12 | 国家电网公司 | A kind of real-time pointer instrument reading method and device based on ARM platforms |
CN108009535A (en) * | 2017-11-21 | 2018-05-08 | 武汉中元华电科技股份有限公司 | A kind of simple pointer meter reading method based on machine vision |
US20200034987A1 (en) * | 2018-07-25 | 2020-01-30 | Beijing Smarter Eye Technology Co. Ltd. | Method and device for building camera imaging model, and automated driving system for vehicle |
CN109558871A (en) * | 2018-10-26 | 2019-04-02 | 中国科学院长春光学精密机械与物理研究所 | A kind of readings of pointer type meters recognition methods and device |
CN110189314A (en) * | 2019-05-28 | 2019-08-30 | 长春大学 | Image positioning method of automobile instrument panel based on machine vision |
CN110414510A (en) * | 2019-07-26 | 2019-11-05 | 华中科技大学 | A method for correcting readings of pointer instruments |
CN110852213A (en) * | 2019-10-30 | 2020-02-28 | 天津大学 | Multi-condition automatic reading method of pointer meter based on template matching |
CN111091076A (en) * | 2019-12-03 | 2020-05-01 | 西北工业大学 | Measurement method of tunnel boundary data based on stereo vision |
CN111080711A (en) * | 2019-12-05 | 2020-04-28 | 东南大学 | A Magnification-Based Calibration Method of Microscopic Imaging System in Nearly Parallel State |
Non-Patent Citations (4)
Title |
---|
LAI, HW等: "A Novel Scale Recognition Method for Pointer Meters Adapted to Different Types and Shapes", 2019 IEEE 15TH INTERNATIONAL CONFERENCE ON AUTOMATION SCIENCE AND ENGINEERING, pages 374 - 379 * |
LI, Q等: "Automatic Reading System based on Automatic Alignment Control for Pointer Meter", IECON 2014 - 40TH ANNUAL CONFERENCE OF THE IEEE INDUSTRIAL ELECTRONICS SOCIETY, pages 3414 - 3418 * |
佘世洲;宋凯;刘辉;谭守标;张骥;: "一种电力指针式仪表示数自动识别的鲁棒方法", 计算机技术与发展, no. 04, pages 200 - 203 * |
金施群;张莉;范利勤;: "单幅图像的三维重构视觉系统误差修正", 应用科学学报, no. 06, pages 81 - 85 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112668578A (en) * | 2020-12-31 | 2021-04-16 | 中广核研究院有限公司 | Pointer instrument reading method and device, computer equipment and storage medium |
CN112668578B (en) * | 2020-12-31 | 2023-11-07 | 中广核研究院有限公司 | Pointer type instrument reading method, pointer type instrument reading device, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110189314B (en) | Image positioning method of automobile instrument panel based on machine vision | |
CN107167169B (en) | Pointer meter reading recognition measurement method based on machine vision system | |
CN111368906B (en) | A reading recognition method of pointer oil level gauge based on deep learning | |
CN110675376A (en) | A PCB defect detection method based on template matching | |
CN104754264A (en) | Feature point based projection deformation correcting method and system | |
CN107506801B (en) | Fish property measuring device and method | |
CN107633533B (en) | High-precision circular marker point center positioning method and device under large distortion lens | |
CN109544628A (en) | A kind of the accurate reading identifying system and method for pointer instrument | |
CN111260720A (en) | A target height determination system based on deep learning method | |
CN110852213B (en) | Pointer instrument multi-condition automatic reading method based on template matching | |
CN111047586B (en) | A Pixel Equivalent Measurement Method Based on Machine Vision | |
CN107578047A (en) | Eccentricity Detection Method of Power Cable | |
CN115187612A (en) | Plane area measuring method, device and system based on machine vision | |
CN104771189A (en) | Three-dimensional head image alignment method and device | |
CN111738971B (en) | Circuit board stereoscopic scanning detection method based on line laser binocular stereoscopic vision | |
CN114323591B (en) | Optical distortion measuring method and testing system based on digital image correlation method | |
CN112001379A (en) | Correction algorithm of automobile instrument fixed viewpoint reading instrument based on machine vision | |
CN116205993A (en) | A high-precision calibration method for bi-telecentric lens for 3D AOI | |
CN112935562A (en) | Laser precision machining method based on paraxial offline measurement | |
CN109470290B (en) | Automatic calibration method and device for instrument pointer | |
CN112990190A (en) | Method and device for identifying reading of pointer instrument | |
CN110864878A (en) | Method for detecting display distortion of high-efficiency large-view-field flat display system | |
CN107945236B (en) | Sub-pixel level nonlinear calibration method suitable for general optical system | |
CN113869112A (en) | A method and device for automatic reading of instruments based on machine vision | |
CN113610086A (en) | Reading correction method and device for vertical-scale pointer instrument and terminal equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
AD01 | Patent right deemed abandoned | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20240927 |