技术领域technical field
本发明涉及视觉测量领域，尤其涉及一种叶片非接触式检测装置及方法。The invention relates to the field of visual measurement, in particular to a blade non-contact detection device and method.
背景技术Background technique
燃气轮机被广泛应用于航空、航天、能源等领域，作为燃气轮机中的核心零件，透平叶片具有结构复杂、制造工艺要求高、测量参数多等特点，其形状误差影响整个燃气轮机的能量转换效率，因此对不规则型面测量的精准性、有效性提出要求。由于透平叶片的尺寸大、种类多、生产制造数量多，如何在保证测量精度符合要求的前提下，针对不同型号的透平叶片，提高测量不同参数的效率有待解决。Gas turbines are widely used in aviation, aerospace, energy and other fields. As the core parts of gas turbines, turbine blades have the characteristics of complex structure, high manufacturing process requirements, and many measurement parameters. Their shape errors affect the energy conversion efficiency of the entire gas turbine. Therefore, It puts forward requirements for the accuracy and effectiveness of irregular surface measurement. Due to the large size, variety and quantity of turbine blades, how to improve the efficiency of measuring different parameters for different types of turbine blades under the premise of ensuring that the measurement accuracy meets the requirements remains to be solved.
目前检测叶片的方式主要分为接触式和非接触式两大类。接触式测量需要使传感器测头直接接触待测叶片，记录接触时的三维空间坐标值，主要代表为专用量具测量法和现在被广泛应用的三坐标测量机。非接触测量主要是基于光学、声学、磁学等原理，通过传感器采集物理模拟量，再通过相应算法转换成工件表面特征参数，主要代表为光学投影法和激光三角法。At present, the methods of detecting blades are mainly divided into two categories: contact and non-contact. Contact measurement requires the sensor probe to directly contact the blade to be measured, and record the three-dimensional coordinate values at the time of contact, which is mainly represented by the special measuring tool measurement method and the three-coordinate measuring machine that is now widely used. Non-contact measurement is mainly based on the principles of optics, acoustics, and magnetism. Physical analog quantities are collected through sensors, and then converted into surface characteristic parameters of the workpiece through corresponding algorithms. The main representatives are optical projection method and laser triangulation method.
上述叶片测量方法都具有自身测量特点。专用量具测量法是指根据叶片的理论型线设计制造且与叶型界面的叶盆叶背相对应的一套母板量具，用叶片固定座及型面测具固定叶片后，用叶盆型面样板和叶背型面样板分别检查叶盆、叶背的型面。这种方法在锻件成型、打孔、粗加工后、人工磨抛前、人工磨抛后等被广泛应用。虽然容易在车间里进行、操作简单，但是人为因素测量误差大、精度较差、效率低、无法测量型面的复杂参数、样板制造周期长及成本高；三坐标测量机测量叶片时，测头将与叶片表面直接接触，会造成测头磨损或叶片表面刮痕。虽然测量精度非常高，但是由于需要测头半径补偿，导致某些小型叶片的型面边缘难以检测。一般的三坐标测量机体积庞大，测量过程繁琐，不适宜现场测量；光学投影法对叶片切面进行投影，测量精度高，但是效率低、对叶片的环境要求高；激光三角法是目前最成熟、应用最为广泛的非接触测量方法，测量速度快、准确度高、结构简单、采样频率高，但是在视场范围内要确保被测物体不被遮挡，激光提取效果也受到不同材质表面反光的影响。综上，现有的叶片测量方法缺陷有：(1)测量效率低；(2)测量精度无法保证；(3)操作复杂；(4)不易于搬运部署且工作环境要求高；(5)测量分析过程繁杂。The above blade measurement methods all have their own measurement characteristics. The special measuring tool measurement method refers to a set of master plate measuring tools designed and manufactured according to the theoretical shape line of the blade and corresponding to the back of the blade pot and leaf of the blade shape interface. Check the profiles of the leaf pot and the leaf back with the surface template and the leaf back profile respectively. This method is widely used after forging forming, drilling, rough machining, before manual grinding and polishing, and after manual grinding and polishing. Although it is easy to carry out in the workshop and the operation is simple, the measurement error due to human factors is large, the accuracy is poor, the efficiency is low, the complex parameters of the profile cannot be measured, the sample manufacturing cycle is long and the cost is high; when the three-coordinate measuring machine measures the blade, the probe Will come into direct contact with the blade surface, causing wear on the probe or scratches on the blade surface. Although the measurement accuracy is very high, the profile edge of some small blades is difficult to detect due to the need for probe radius compensation. The general three-coordinate measuring machine is bulky, the measurement process is cumbersome, and it is not suitable for on-site measurement; the optical projection method projects the cutting surface of the blade, and the measurement accuracy is high, but the efficiency is low and the environmental requirements for the blade are high; the laser triangulation method is currently the most mature, The most widely used non-contact measurement method has fast measurement speed, high accuracy, simple structure, and high sampling frequency. However, it must be ensured that the measured object is not blocked within the field of view, and the laser extraction effect is also affected by the reflection of the surface of different materials. . In summary, the existing blade measurement methods have the following defects: (1) low measurement efficiency; (2) measurement accuracy cannot be guaranteed; (3) complex operation; (4) difficult to transport and deploy and require high working environment; (5) measurement The analysis process is complicated.
发明内容Contents of the invention
本发明的目的在于通过一种叶片非接触式检测装置及方法，来解决以上背景技术部分提到的问题。The object of the present invention is to solve the problems mentioned above in the background technology section through a blade non-contact detection device and method.
为达此目的，本发明采用以下技术方案：For reaching this purpose, the present invention adopts following technical scheme:
一种叶片非接触式检测装置，该装置包括线性模组、视觉传感器、转台以及控制柜；所述视觉传感器安装在所述线性模组上；待测叶片安装在转台上；所述控制柜与线性模组、视觉传感器、转台通讯连接，用于控制线性模组和转台运动，使视觉传感器采集所述待测叶片的完整叶片点云数据，并对所述完整叶片点云数据进行处理，输出所述待测叶片的检测结果。A blade non-contact detection device, the device includes a linear module, a visual sensor, a turntable and a control cabinet; the visual sensor is installed on the linear module; the blade to be tested is installed on the turntable; the control cabinet and The linear module, the vision sensor, and the turntable communication connection are used to control the movement of the linear module and the turntable, so that the vision sensor collects the complete blade point cloud data of the blade to be tested, and processes the complete blade point cloud data to output The detection result of the blade to be tested.
特别地，所述控制线性模组和转台运动，使视觉传感器采集所述待测叶片的完整叶片点云数据，具体包括：控制线性模组和转台实现三轴联动，使视觉传感器采集所述待测叶片的完整叶片点云数据。In particular, the controlling the motion of the linear module and the turntable so that the visual sensor collects the complete blade point cloud data of the blade to be tested includes: controlling the linear module and the turntable to realize three-axis linkage, so that the visual sensor collects the blade point cloud data of the blade to be tested. The complete blade point cloud data of the measured blade.
特别地，所述对所述完整叶片点云数据进行处理，输出所述待测叶片的检测结果，具体包括：对所述完整叶片点云数据进行处理，输出可视化、定制化的叶片分析报告。In particular, the processing the point cloud data of the complete blade and outputting the detection result of the blade to be tested specifically includes: processing the point cloud data of the complete blade and outputting a visualized and customized blade analysis report.
特别地，根据待测叶片不同的测量任务调整视觉传感器的类型及其安装时与线性模组在竖直方向的角度；所述视觉传感器在指定运动位置被触发，根据不同的测量任务调整不同的点云密度并在指定区域进行扫描检测。In particular, adjust the type of the vision sensor and the vertical angle between the linear module and the linear module according to the different measurement tasks of the blade to be tested; the vision sensor is triggered at the specified movement position, and adjust different Point cloud density and scanning detection in the specified area.
特别地，所述线性模组的数量设置为两个，每个线性模组上安装一个所述视觉传感器。In particular, the number of the linear modules is set to two, and one visual sensor is installed on each linear module.
特别地，所述视觉传感器采用通用的视觉标定板标定相机参数、激光光刀平面位姿、视觉传感器运动方向。In particular, the visual sensor adopts a common visual calibration board to calibrate the camera parameters, the plane pose of the laser light knife, and the motion direction of the visual sensor.
特别地，所述视觉传感器的坐标系统一方式如下：在扫描空间内，分别用两个视觉传感器扫描同一个标定物，获取标定物的位姿，通过刚体变换的方法求出两个视觉传感器之间的坐标系转换关系；通过旋转转台使视觉传感器多个角度扫描标定物的位姿，建立系统坐标系。In particular, the first mode of the coordinate system of the visual sensor is as follows: in the scanning space, use two visual sensors to scan the same calibration object to obtain the pose of the calibration object, and obtain the distance between the two visual sensors through the method of rigid body transformation. The coordinate system conversion relationship among them; by rotating the turntable, the visual sensor scans the pose of the calibration object at multiple angles to establish a system coordinate system.
特别地，所述使视觉传感器采集所述待测叶片的完整叶片点云数据，并对所述完整叶片点云数据进行处理，输出所述待测叶片的检测结果，具体包括：将两个视觉传感器每次扫描的点云数据按照刚体变换的方法粗配准到同一个系统坐标系下，通过迭代最近点(Iterative Closest Point，ICP)方法对粗配准的点云进行精配准，并将精配准后的点云进行显示，然后，把所述精配准后的点云与相应的数字化样板进行比对，输出可视化、定制化的叶片分析报告。In particular, the making the visual sensor collect the complete blade point cloud data of the blade to be tested, and process the complete blade point cloud data, and output the detection result of the blade to be tested, specifically includes: combining two visual The point cloud data of each scan of the sensor is roughly registered to the same system coordinate system according to the rigid body transformation method, and the coarsely registered point cloud is finely registered by the Iterative Closest Point (ICP) method, and the The point cloud after fine registration is displayed, and then the point cloud after fine registration is compared with the corresponding digital template, and a visualized and customized blade analysis report is output.
特别地，所述叶片非接触式检测装置还包括工作平台和滚轮；所述转台设置在工作平台上；所述滚轮安装在工作平台底部；所述两个线性模组对称排布在工作平台上。In particular, the blade non-contact detection device also includes a working platform and rollers; the turntable is set on the working platform; the rollers are installed at the bottom of the working platform; the two linear modules are symmetrically arranged on the working platform .
本发明还公开了一种应用上述叶片非接触式检测装置的叶片非接触式检测方法，该方法包括：The present invention also discloses a blade non-contact detection method using the above blade non-contact detection device, the method comprising:
S101、将待测叶片装夹在所述叶片非接触式检测装置的转台上；S101, clamping the blade to be tested on the turntable of the blade non-contact detection device;
S102、控制柜控制线性模组和转台实现三轴联动，使两个视觉传感器进行第一次向上扫描，随后转台旋转指定角度，两个视觉传感器进行第一次向下扫描，按照以上扫描方式循环扫描直至采集到所述待测叶片的完整叶片点云数据；S102. The control cabinet controls the linear module and the turntable to achieve three-axis linkage, so that the two visual sensors scan upward for the first time, and then the turntable rotates at a specified angle, and the two visual sensors scan downward for the first time, and cycle according to the above scanning method Scanning until the complete blade point cloud data of the blade to be tested is collected;
S103、控制柜将两个视觉传感器每次扫描的点云数据按照刚体变换的方法粗配准到同一个系统坐标系下，通过迭代最近点(Iterative Closest Point，ICP)方法对粗配准的点云进行精配准，并将精配准后的点云进行显示；S103. The control cabinet roughly registers the point cloud data of each scan of the two visual sensors into the same system coordinate system according to the method of rigid body transformation, and uses the Iterative Closest Point (ICP) method to roughly register the points The cloud is finely aligned, and the finely aligned point cloud is displayed;
S104、控制柜把所述精配准后的点云与相应的数字化样板进行比对，输出可视化、定制化的叶片分析报告。S104. The control cabinet compares the finely aligned point cloud with the corresponding digital template, and outputs a visualized and customized blade analysis report.
本发明提出的叶片非接触式检测装置及方法优点如下：一、操作简单；操作人员只需将叶片装夹到工作平台上的转台上，在控制柜的软件中进行叶片型号、测量项目等相关设置后，启动测量程序即可等待分析结果，简化检测流程降低了操作人员的操作难度；二、效率高；视觉传感器根据非接触式激光三角测量原理，配合线性模组以及转台的运动，对视场内的待测叶片进行多角度全面扫描，由于视觉传感器是由指定扫描位置触发扫描，整个扫描过程中分为向上向下两个方向，再一次提高了装置的整体扫描效率；三、自动化程度高；在测量过程，操作人员除了将叶片装夹到工作台上，只需要在软件中设置相应参数并启动测量即可，非接触式检测装置就会自动进行快速检测，并且输出相应检测分析报告，实现了检测分析一体化、自动化。The advantages of the blade non-contact detection device and method proposed by the present invention are as follows: 1. The operation is simple; the operator only needs to clamp the blade to the turntable on the working platform, and perform the blade model, measurement items, etc. in the software of the control cabinet. After setting, start the measurement program and wait for the analysis results, simplifying the detection process and reducing the difficulty of the operator's operation; 2. High efficiency; the visual sensor is based on the principle of non-contact laser triangulation, and cooperates with the linear module and the movement of the turntable. The blades to be tested in the field are fully scanned from multiple angles. Since the visual sensor is triggered by the designated scanning position, the entire scanning process is divided into two directions, up and down, which once again improves the overall scanning efficiency of the device; 3. Degree of automation High; during the measurement process, the operator only needs to set the corresponding parameters in the software and start the measurement in addition to clamping the blade on the workbench. The non-contact detection device will automatically perform rapid detection and output a corresponding detection analysis report , Realized the integration and automation of detection and analysis.
附图说明Description of drawings
图1为本发明实施例提供的叶片非接触式检测装置结构图；Fig. 1 is a structural diagram of a blade non-contact detection device provided by an embodiment of the present invention;
图2A和图2B为本发明实施例提供的视觉传感器的原理简图；Fig. 2A and Fig. 2B are schematic diagrams of the principle of the vision sensor provided by the embodiment of the present invention;
图3A和图3B为本发明实施例提供的各坐标系示意图；3A and 3B are schematic diagrams of coordinate systems provided by embodiments of the present invention;
图4为本发明实施例提供的误差伪彩图；Fig. 4 is the error pseudo-color map provided by the embodiment of the present invention;
图5为本发明实施例提供的定制化叶片截面分析报告；Fig. 5 is the customized blade section analysis report provided by the embodiment of the present invention;
图6为本发明实施例提供的点云数据处理流程示意图；6 is a schematic diagram of a point cloud data processing flow provided by an embodiment of the present invention;
图7为本发明实施例提供的叶片非接触式检测方法流程图。Fig. 7 is a flow chart of a blade non-contact detection method provided by an embodiment of the present invention.
具体实施方式Detailed ways
为了便于理解本发明，下面将参照相关附图对本发明进行更全面的描述。附图中给出了本发明的较佳实施例。但是，本发明可以以许多不同的形式来实现，并不限于本文所描述的实施例。相反地，提供这些实施例的目的是使对本发明的公开内容理解的更加透彻全面。需要说明的是，当一个元件被认为是“连接”另一个元件，它可以是直接连接到另一个元件或者可能同时存在居中元件。除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。本文所使用的术语“及/或”包括一个或多个相关的所列项目的任意的和所有的组合。In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the associated drawings. Preferred embodiments of the invention are shown in the accompanying drawings. However, the present invention can be embodied in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. It should be noted that when an element is considered to be "connected" to another element, it may be directly connected to the other element or there may be intervening elements at the same time. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field of the invention. The terms used herein in the description of the present invention are for the purpose of describing specific embodiments only, and are not intended to limit the present invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
请参照图1所示，图1为本发明实施例提供的叶片非接触式检测装置结构图。本实施例中叶片非接触式检测装置包括线性模组101、视觉传感器102、转台103以及控制柜104。所述视觉传感器102安装在所述线性模组101上。待测叶片安装在转台103上。所述控制柜104与线性模组101、视觉传感器102、转台103通讯连接，用于控制线性模组101和转台103运动，使视觉传感器102采集所述待测叶片的完整叶片点云数据，并对所述完整叶片点云数据进行处理，输出所述待测叶片的检测结果。Please refer to FIG. 1 , which is a structural diagram of a blade non-contact detection device provided by an embodiment of the present invention. The blade non-contact detection device in this embodiment includes a linear module 101 , a vision sensor 102 , a turntable 103 and a control cabinet 104 . The vision sensor 102 is installed on the linear module 101 . The blade to be tested is installed on the turntable 103 . The control cabinet 104 communicates with the linear module 101, the visual sensor 102, and the turntable 103, and is used to control the movement of the linear module 101 and the turntable 103, so that the visual sensor 102 collects the complete blade point cloud data of the blade to be tested, and The point cloud data of the complete blade is processed, and the detection result of the blade to be measured is output.
具体的，在本实施例中所述线性模组101的数量设置为两个，每个线性模组101上安装一个所述视觉传感器102。为了使叶片非接触式检测装置能够灵活移动，在本实施例中所述叶片非接触式检测装置还设置有工作平台105和滚轮106；所述转台103设置在工作平台105上；所述滚轮106安装在工作平台105底部；所述两个线性模组101对称排布在工作平台105上。Specifically, in this embodiment, the number of linear modules 101 is set to two, and one visual sensor 102 is installed on each linear module 101 . In order to enable the blade non-contact detection device to move flexibly, the blade non-contact detection device is also provided with a working platform 105 and rollers 106 in this embodiment; the turntable 103 is arranged on the working platform 105; the rollers 106 It is installed at the bottom of the working platform 105; the two linear modules 101 are symmetrically arranged on the working platform 105.
在本实施例中所述待测叶片在转台103上的装夹方式为竖直立式抓夹，由于非接触式视觉测量的限制，叶片测量部分及周边不能被遮挡，装夹位置应在叶根与叶片顶部，叶片装夹测量过程中应保持相对转台103静止，旋转过程中叶片应尽量避免振动，其中，待测叶片具体装夹位置应与具体叶片型号的装夹要求一致，普遍利用叶片上下打的基准孔进行定位装夹，为了避免选择过程中叶片振动，针对上下打基准孔进行定位装夹的夹方式，设计的夹具应保证比较高的装配定位精度，在不遮挡叶片的前提下，上顶孔的顶针应与转盘随动。根据不同生产工序的特点，叶片夹具需要针对其特征设计，例如叶片锻压件等还没有测量基准时，可以只是保证叶片能够被平稳可靠加紧即可，而如果是粗加工件等已有测量基准时，叶片夹具既要求叶片按照基准平稳可靠装夹，又要保证基准在视觉传感器102视场范围内。对于不同类型、尺寸的叶片，叶片夹具应尽量满足相同工序叶片装夹的通用性、稳定性、简易性。In this embodiment, the clamping method of the blade to be measured on the turntable 103 is vertical clamping. Due to the limitation of non-contact visual measurement, the blade measurement part and its surroundings cannot be blocked, and the clamping position should be on the blade. The root and the top of the blade should remain stationary relative to the turntable 103 during the blade clamping measurement process, and the blade should avoid vibration as much as possible during the rotation process. Among them, the specific clamping position of the blade to be tested should be consistent with the clamping requirements of the specific blade model. The reference holes drilled up and down are used for positioning and clamping. In order to avoid blade vibration during the selection process, the clamping method for positioning and clamping of the reference holes drilled up and down should be designed to ensure relatively high assembly and positioning accuracy. On the premise of not blocking the blades , the thimble of the upper hole should follow the turntable. According to the characteristics of different production processes, blade fixtures need to be designed according to their characteristics. For example, when there is no measurement benchmark for blade forgings, it can just ensure that the blade can be tightened smoothly and reliably. , the blade fixture not only requires the blade to be stably and reliably clamped according to the reference, but also ensures that the reference is within the field of view of the visual sensor 102 . For blades of different types and sizes, the blade fixture should try to meet the versatility, stability and simplicity of blade clamping in the same process.
在本实施例中所述视觉传感器102采用CCD相机201(下文简称相机)和线激光器202(下文简称激光器)配置，其原理结构简图如图2A和图2B所示，基于激光三角测量原理，根据已知的线激光宽度W与角度β可计算相机光心到待测物体的距离即视场深度，相机光心到激光平面203(激光平面即指光刀平面)的距离B、相机光轴与基线夹角由于基线垂直于光刀平面为已知的视觉传感器102参数，由此可以根据三角形余弦或正弦定理计算出激光器到待测物体的距离D。视觉传感器102的测量坐标系建立在光刀平面上，光心与光刀平面的垂足作为测量坐标系的原点，该点指向光心的方向为y轴，相机光轴在光刀平面的投影方向为z轴，x轴为两者的叉乘。线激光用于提供测量空间中的约束平面，CCD相机201用于获取激光条纹的图像。通过激光条纹中心提取算法对图像中受场景调制后的激光条纹中心进行提取，获得激光条纹的中心在图像坐标系下的坐标，从而实现对激光光条的三维重建。其中，所述图像坐标系是指二维图像上，以图像内某一点为原点，X轴与Y轴与像素坐标系中u,v轴平行的二维直角坐标系，单位为物理单位mm，像素坐标系是以图像左上角为原点，u轴平行于图片长方向向右，v轴平行于图片宽方向像下，单位为像素。图像坐标系与相机坐标系的关系是通过相机标定得出的一个内参矩阵。In the present embodiment, the visual sensor 102 adopts a CCD camera 201 (hereinafter referred to as a camera) and a line laser 202 (hereinafter referred to as a laser) configuration, and its schematic structural diagram is shown in Figure 2A and Figure 2B, based on the principle of laser triangulation, According to the known line laser width W and angle β, the distance from the optical center of the camera to the object to be measured can be calculated That is, the depth of field of view, the distance B from the optical center of the camera to the laser plane 203 (the laser plane refers to the plane of the light knife), and the angle between the optical axis of the camera and the baseline Since the baseline perpendicular to the light knife plane is a known parameter of the vision sensor 102, the distance D from the laser to the object to be measured can be calculated according to the triangular cosine or sine law. The measurement coordinate system of the visual sensor 102 is established on the light knife plane, and the vertical foot of the optical center and the light knife plane is used as the origin of the measurement coordinate system. The direction pointing to the optical center is the y-axis, and the projection direction of the camera optical axis on the light knife plane is z axis, and the x-axis is the cross product of the two. The line laser is used to provide a constrained plane in the measurement space, and the CCD camera 201 is used to acquire images of the laser stripes. The laser stripe center in the image modulated by the scene is extracted by the laser stripe center extraction algorithm, and the coordinates of the laser stripe center in the image coordinate system are obtained, so as to realize the three-dimensional reconstruction of the laser light stripe. Wherein, the image coordinate system refers to a two-dimensional rectangular coordinate system on a two-dimensional image, with a certain point in the image as the origin, and the X-axis and Y-axis are parallel to the u and v axes in the pixel coordinate system, and the unit is the physical unit mm, The pixel coordinate system is based on the upper left corner of the image as the origin, the u-axis is parallel to the long direction of the image to the right, and the v-axis is parallel to the bottom of the image's width direction, and the unit is pixel. The relationship between the image coordinate system and the camera coordinate system is an internal reference matrix obtained through camera calibration.
在本实施例中所述叶片非接触式检测装置各坐标系建立示意图如图3A和图3B所示，图中Pc(Xc,Yc,Zc)指相机坐标系，PI(XI,YI,ZI)指激光平面坐标系，Ps(Xs,Ys,Zs)指系统坐标系。激光平面坐标系Y轴竖直向上，Z轴分别指向转盘中心，X轴为两者的叉乘，由于两个视觉传感器102的坐标系呈对称分布，两个视觉传感器102激光平面坐标系的X轴方向相反，需要说明的是，若两个视觉传感器102安装角度相同，例如都为竖直安装，那么其中一个视觉传感器将只是增加测量效率；若两个视觉传感器102安装角度不同，例如一个视场向上，一个视场向下，则扫描测量的区域不同；对于叶片测量的具体需求而确定安装角度，只用一个视觉传感器102不能满足所有的测量需求。根据上述视觉传感器102原理，相机坐标系在激光平面坐标系的基础上向上平移了距离B，并且向转台103方向旋转θ。系统坐标系的原点为转台103中心，Z轴为转台103平面法向向上，X轴为Z轴与相机光轴方向的叉乘，Y轴为Z轴与X轴的叉乘。相机拍摄照片，通过激光条纹中心提取算法可以获得图片上光条中心点的二维坐标，通过相机标定，可将这些点转化到相机坐标系下表示，再通过激光平面标定后，将点云坐标转换到了激光平面坐标系下，第二个视觉传感器在激光平面坐标系下的点云坐标通过设备位姿关系转换到第一个视觉传感器102的激光平面坐标系下，最后将第一个视觉传感器102的激光平面坐标系下的点云坐标转换到转台坐标系即系统坐标系下，需要说明的是，第一个视觉传感器和第二个视觉传感器只是为了便于阐述而进行的人为定义，两个视觉传感器102中若一个为第一个视觉传感器，那另一个自然就为第二个视觉传感器。针对上述设备位姿关系解释如下：由于两个视觉传感器102扫描后对图片进行激光条纹提取，将各自的点都转化到各自的激光平面坐标系下，设备位姿关系就是两个视觉传感器102之间的坐标系转换关系(详见下文具体说明)，将一个视觉传感器扫描出的所有点云转化到另一个视觉传感器的激光平面坐标系下表示。In this embodiment, the schematic diagrams for establishing each coordinate systemof the blade non-contactdetection device are shown in FIG. 3A andFIG .I , YI , ZI ) refers to the laser plane coordinate system, and Ps (Xs , Ys , Zs ) refers to the system coordinate system. The Y-axis of the laser plane coordinate system is vertically upward, the Z-axis points to the center of the turntable respectively, and the X-axis is the cross product of the two. Since the coordinate systems of the two visual sensors 102 are symmetrically distributed, the X axis of the laser plane coordinate system of the two visual sensors 102 The directions of the axes are opposite. It should be noted that if the two vision sensors 102 are installed at the same angle, for example, they are installed vertically, then one of the vision sensors will only increase the measurement efficiency; if the two vision sensors 102 are installed at different angles, such as one If the field of view is upward and the field of view is downward, the scanning measurement area is different; the installation angle is determined according to the specific requirements of blade measurement, and only one vision sensor 102 cannot meet all measurement requirements. According to the principle of the above-mentioned visual sensor 102 , the camera coordinate system is translated upward by a distance B on the basis of the laser plane coordinate system, and rotates θ toward the direction of the turntable 103 . The origin of the system coordinate system is the center of the turntable 103, the Z axis is the normal direction of the turntable 103 plane upward, the X axis is the cross product of the Z axis and the optical axis direction of the camera, and the Y axis is the cross product of the Z axis and the X axis. The camera takes pictures, and the two-dimensional coordinates of the center point of the light stripe on the picture can be obtained through the laser stripe center extraction algorithm. Through camera calibration, these points can be transformed into the camera coordinate system for representation, and then the point cloud coordinates can be obtained after laser plane calibration Converted to the laser plane coordinate system, the point cloud coordinates of the second visual sensor in the laser plane coordinate system are converted to the laser plane coordinate system of the first visual sensor 102 through the device pose relationship, and finally the first visual sensor The point cloud coordinates in the laser plane coordinate system of 102 are transformed into the turntable coordinate system, that is, the system coordinate system. It should be noted that the first visual sensor and the second visual sensor are only artificially defined for the convenience of explanation. The two If one of the vision sensors 102 is the first vision sensor, the other is naturally the second vision sensor. The above-mentioned equipment pose relationship is explained as follows: Since the two visual sensors 102 scan the image and extract the laser stripes, and transform the respective points into their respective laser plane coordinate systems, the equipment pose relationship is the distance between the two visual sensors 102. The coordinate system conversion relationship between them (see below for details), transform all the point clouds scanned by one visual sensor into the laser plane coordinate system of another visual sensor.
在本实施例中，当第一次使用本发明时，需要做整个叶片非接触式检测装置的系统标定，该标定工作只需要做一次或后续测量中需要重新校准。整个标定过程只需要一个通用的视觉标定板(如简易平面标定板)以及一个标定物(如标定小球)，所述视觉传感器102采用通用的视觉标定板标定相机参数、激光光刀平面位姿、视觉传感器102运动方向，而不需要定制特殊形状的标定块，可大大简化标定过程和标定成本。所述两个视觉传感器102坐标系的统一方式如下：在扫描空间内，分别用两个视觉传感器102扫描同一个标定物，获取标定物的位姿，通过刚体变换的方法求出两个视觉传感器102之间的坐标系转换关系；通过旋转转台103使视觉传感器102多个角度扫描标定物的位姿，建立系统坐标系。以上标定完成后，通过三角测量原理获取单幅点云，两个视觉传感器102扫描出的点云数据可以直接粗配准在同一坐标系下。In this embodiment, when the present invention is used for the first time, system calibration of the non-contact detection device for the entire blade needs to be done, and the calibration work only needs to be done once or needs to be re-calibrated in subsequent measurements. The entire calibration process only needs a common visual calibration board (such as a simple plane calibration board) and a calibration object (such as a calibration ball), and the vision sensor 102 uses a general visual calibration board to calibrate camera parameters, laser light knife plane pose, The movement direction of the vision sensor 102 does not need to customize a calibration block with a special shape, which greatly simplifies the calibration process and calibration cost. The unification of the coordinate systems of the two visual sensors 102 is as follows: in the scanning space, use two visual sensors 102 to scan the same calibration object respectively to obtain the pose of the calibration object, and obtain the coordinates of the two visual sensors through the method of rigid body transformation. 102 coordinate system conversion relationship; by rotating the turntable 103, the visual sensor 102 scans the pose of the calibration object at multiple angles to establish a system coordinate system. After the above calibration is completed, a single point cloud is acquired through the principle of triangulation, and the point cloud data scanned by the two visual sensors 102 can be directly roughly registered in the same coordinate system.
由于本实施例提出的叶片非接触式检测装置的检测原理是利用扫描的叶片点云与数字化样板进行比对，因此在进行检测之前还需要获得不同型号叶片的CAD模型，并且根据不同的测量任务设置不同的数字化样板。随后在每次的检测中，用户只需要在交互软件中选择相应的数字化样板，设置好相应的参数例如相机曝光时间、相机拍摄帧率、起始扫描位置、扫描间隔、扫描次数等，就可以对叶片进行自动化检测分析。以普通的叶片指定截面作测量分析为例，制作数字化样板过程为在Polyworks软件中导入叶片三维模型，在需要作截面分析的高度分别做截面特征。每次测量用宏命令完成以下分析过程：导入点云，根据测量基准将点云配准到CAD理论模型上，如果叶片为锻件，没有测量基准，则进行全局拟合，测量每个截面的最大厚度、弦长、背弧、内弧等参数，如图4所示生成叶片表面误差伪彩图，调整误差伪彩图的误差范围，如图5所示将每个截面的参数、截图、叶片表面伪彩图输出到报告中，导出报告后删除当前所有测量项目。Since the detection principle of the blade non-contact detection device proposed in this embodiment is to compare the scanned blade point cloud with the digital template, it is necessary to obtain CAD models of different types of blades before detection, and according to different measurement tasks Set up different digitization templates. Then in each inspection, the user only needs to select the corresponding digital template in the interactive software, set the corresponding parameters such as camera exposure time, camera shooting frame rate, initial scanning position, scanning interval, scanning times, etc. Automated detection and analysis of leaves. Taking the measurement and analysis of a specified section of an ordinary blade as an example, the process of making a digital template is to import the 3D model of the blade into the Polyworks software, and make section features at the heights that need to be analyzed. Use macro commands for each measurement to complete the following analysis process: import point cloud, register the point cloud to the CAD theoretical model according to the measurement benchmark, if the blade is a forging and there is no measurement benchmark, perform global fitting, measure the maximum Thickness, chord length, back arc, inner arc and other parameters, as shown in Figure 4, generate a pseudo-color map of the error of the blade surface, adjust the error range of the error pseudo-color map, as shown in Figure 5, the parameters of each section, screenshots, blades The surface pseudo-color map is output to the report, and all current measurement items are deleted after the report is exported.
所述控制柜104应集成叶片非接触式检测装置的标定算法、激光条纹中心提取算法、点云配准算法、点云定制化分析算法等，利用了PCI-E接口技术，将线性模组101控制模块和点云处理模块集成在一起，能够同时控制线性模组101、转台103的运动与处理视觉传感器102采集的测量数据的能力，不仅可以获取叶片完整点云，同时可直接对点云数据进行处理、分析及测量，自动快速输出分析结果，分析结果为含有表格、图片、自定义参数的报告。下面对上述标定算法、激光条纹中心提取算法、点云配准算法、点云定制化分析算法扼要介绍如下：1.标定算法包括：相机标定算法、激光平面标定算法、运动方向标定算法、多扫描头位姿关系标定算法以及转台标定算法。其中相机标定算法用来标定相机内参；激光平面标定算法用来确定激光平面坐标系与相机坐标系的关系；运动方向标定算法用来确定实际运动方向向量；多扫描头位姿关系标定算法用来进行两个视觉传感器之间的坐标系转换；转台标定算法用来求转盘旋转轴以及视觉传感器与转台坐标系之间的变换关系。2.激光条纹中心提取算法：将激光建模为矩形的台阶函数，估计激光图片的背景亮度和前景亮度，确定亮条纹宽度，再将窗口内的有效像素参与重心计算，得到激光条纹重心。3.点云配准算法：先将所有点云按照标定的坐标系转换关系，全部转换到一个坐标系中粗配准，再用ICP最近迭代算法进行精配准。4.点云定制化分析算法：需根据具体的叶片测量要求：例如叶片型面指定点测量是将指定点都在CAD中建立好点特征，然后每次导入测量的点云与CAD配准，计算每个点特征到最近点云的法向距离，自动生成所有距离报告；例如叶片截面分析是先导入叶片CAD和点云，进行配准，建立相应截面的叶片量规，测量每个叶片截面参数，如最大厚度、弦宽、前缘、后缘等，自动生成配准的伪彩图以及参数报告；例如叶片锻件错模分析是先导入叶片上下模具的CAD，在CAD上建立相应的平面、点特征，导入叶片点云后，以上模具为基准配准，再以下模具为基准配准，计算点云在两次配准中的运动变换矩阵，最后令下模具按照该矩阵移动，比较平面、点特征平移和旋转的距离和角度，自动生成模具错移报告。The control cabinet 104 should integrate the calibration algorithm of the blade non-contact detection device, the laser stripe center extraction algorithm, the point cloud registration algorithm, the point cloud customization analysis algorithm, etc., and the PCI-E interface technology is used to integrate the linear module 101 The control module and the point cloud processing module are integrated together, which can simultaneously control the motion of the linear module 101 and the turntable 103 and the ability to process the measurement data collected by the visual sensor 102. Not only can the complete point cloud of the blade be obtained, but also the point cloud data can be directly processed. Perform processing, analysis and measurement, automatically and quickly output the analysis results, and the analysis results are reports containing tables, pictures, and custom parameters. The above calibration algorithm, laser stripe center extraction algorithm, point cloud registration algorithm, and point cloud customization analysis algorithm are briefly introduced as follows: 1. Calibration algorithms include: camera calibration algorithm, laser plane calibration algorithm, motion direction calibration algorithm, multi- Calibration algorithm for scanning head pose relationship and calibration algorithm for turntable. The camera calibration algorithm is used to calibrate the internal parameters of the camera; the laser plane calibration algorithm is used to determine the relationship between the laser plane coordinate system and the camera coordinate system; the motion direction calibration algorithm is used to determine the actual motion direction vector; the multi-scan head pose relationship calibration algorithm is used to The coordinate system conversion between the two vision sensors is performed; the turntable calibration algorithm is used to find the transformation relationship between the rotation axis of the turntable and the coordinate system of the vision sensor and the turntable. 2. Laser stripe center extraction algorithm: Model the laser as a rectangular step function, estimate the background brightness and foreground brightness of the laser image, determine the width of the bright stripe, and then participate in the center of gravity calculation of the effective pixels in the window to obtain the center of gravity of the laser stripe. 3. Point cloud registration algorithm: first convert all point clouds into a coordinate system for rough registration according to the calibrated coordinate system conversion relationship, and then use the ICP nearest iteration algorithm for fine registration. 4. Point cloud customized analysis algorithm: It needs to be based on specific blade measurement requirements: for example, the measurement of specified points on the blade surface is to establish the point features of the specified points in CAD, and then import the measured point cloud and CAD registration each time. Calculate the normal distance from each point feature to the nearest point cloud, and automatically generate all distance reports; for example, for blade section analysis, first import blade CAD and point cloud, perform registration, establish a blade gauge for the corresponding section, and measure the parameters of each blade section , such as the maximum thickness, chord width, leading edge, trailing edge, etc., automatically generate a pseudo-color map and parameter report for registration; for example, the wrong mode analysis of a blade forging is to first import the CAD of the upper and lower molds of the blade, and establish the corresponding plane, Point features, after importing the blade point cloud, the above mold is used as the reference registration, and then the lower mold is used as the reference registration, and the motion transformation matrix of the point cloud in the two registrations is calculated, and finally the lower mold is ordered to move according to the matrix, and the plane, The distance and angle of translation and rotation of point features, and automatic generation of mold misalignment reports.
在本实施例中所述线性模组101应为精密级且具备封闭外壳，行程要大于需要测量的叶片，以保证待测叶片能够被覆盖在视觉传感器102的视场下，对称排布在工作平台105上，中间测量空间为(1000×400×250)mm，线性模组101、转台103的编码器、驱动器、限位开关等接线应封装在测量范围以外。所述转台103、工作平台105、滚轮106应能承受装置其他部件的重量以及待测叶片重量。其中，滚轮106应具备万向移动及卡死功能，搬移设备时利于转移，在进行检测工作时能保证整个装置静止于地面。转台103承重规格为200kg，端跳精度为10μm，绝对定位精度为0.005°。运动控制采用具有四轴控制、位置比较功能的运动控制卡。线性模组101的有效行程为1m，重复精度为±0.005mm。所述视觉传感器102精度要高于叶片测量精度要求(在本实施例中叶片测量要求误差为0.1mm那么视觉传感器的测量精度要小于0.1mm)，如小于0.05mm，测量深度(整个视觉传感器的深度，这里指图2A中的D)为200mm至400mm，CCD相机应具有较高的拍摄帧率与较高的分辨率，选用规格为149fps与1280×1024，基线距离(图2A中的B)为240mm,成像角(图2A中的θ)38.7°，重量为3.5kg，镜头选用焦距为8.0mm，分辨率为120.00lp/mm，畸变率为0.60％，线激光器功率为100mw，发散角(激光器发射出的线激光的角度)60°，根据测量所需，用四个螺钉倾斜或竖直固定在线性模组101上。针对不同叶片特征需要，可以适当调整视觉传感器102于线性模组101在竖直方向上的角度或更换不同视觉传感器102来获取各个角度的点云信息。由于视觉传感器102在指定运动位置被触发，可以根据不同测量任务调整不同的点云密度并且按照指定区域进行扫描检测。In this embodiment, the linear module 101 should be precision-grade and equipped with a closed casing, and the stroke should be larger than the blade to be measured, so as to ensure that the blade to be measured can be covered under the field of view of the visual sensor 102 and symmetrically arranged in the working area. On the platform 105, the intermediate measurement space is (1000×400×250) mm, and the wiring of the linear module 101 and the rotary table 103 such as encoders, drivers, and limit switches should be packaged outside the measurement range. The turntable 103, working platform 105, and rollers 106 should be able to bear the weight of other parts of the device and the weight of the blade to be tested. Among them, the roller 106 should have the function of universal movement and locking, which is convenient for transferring when moving the equipment, and can ensure that the whole device is still on the ground when performing detection work. The load-bearing specification of the turntable 103 is 200kg, the end jump accuracy is 10μm, and the absolute positioning accuracy is 0.005°. Motion control adopts a motion control card with four-axis control and position comparison functions. The effective stroke of the linear module 101 is 1m, and the repeatability is ±0.005mm. Described vision sensor 102 precision will be higher than blade measurement accuracy requirement (in the present embodiment blade measurement requirement error is 0.1mm so the measurement accuracy of vision sensor will be less than 0.1mm), as less than 0.05mm, measuring depth (whole vision sensor Depth, here refers to D) in Figure 2A is 200mm to 400mm, the CCD camera should have a higher shooting frame rate and higher resolution, the selected specifications are 149fps and 1280×1024, the baseline distance (B in Figure 2A) 240mm, the imaging angle (theta in Figure 2A) is 38.7°, the weight is 3.5kg, the focal length of the lens is 8.0mm, the resolution is 120.00lp/mm, the distortion rate is 0.60%, the power of the line laser is 100mw, and the divergence angle ( The angle of the line laser emitted by the laser) is 60°, and it is fixed on the linear module 101 obliquely or vertically with four screws according to the measurement requirements. According to the requirements of different blade characteristics, the vertical angle of the visual sensor 102 to the linear module 101 can be appropriately adjusted or replaced with a different visual sensor 102 to obtain point cloud information at various angles. Since the vision sensor 102 is triggered at a designated moving position, different point cloud densities can be adjusted according to different measurement tasks and scanning detection can be performed according to designated areas.
在所有准备工作完成后，本发明在工作时，所述控制柜104控制线性模组101和转台103运动，使视觉传感器102采集所述待测叶片的完整叶片点云数据，并对所述完整叶片点云数据进行处理，输出所述待测叶片的检测结果。需要说明的是，根据待测叶片不同的测量任务调整视觉传感器102的类型及其安装时与线性模组101在竖直方向的角度；所述视觉传感器102在指定运动位置被触发，根据不同的测量任务调整不同的点云密度并在指定区域进行扫描检测。另外，根据不同叶片的大小和材料反光特性，在每种叶片第一次测量时可能需要进行曝光、运动位置调试以便达到测量最佳效果，例如对同种叶片批次的第一件测量开始前，先将两个测头移动，使激光照射在叶片表面即可，在软件中调整合适的曝光使拍摄的图片适合激光条纹中心提取，运动位置调试是为了确认设置的测量范围是否能够覆盖整个叶片，随后即可开始测量。After all the preparatory work is completed, when the present invention is working, the control cabinet 104 controls the movement of the linear module 101 and the turntable 103, so that the visual sensor 102 collects the complete blade point cloud data of the blade to be tested, and performs The point cloud data of the blade is processed, and the detection result of the blade to be measured is output. It should be noted that the type of the visual sensor 102 and the vertical angle between the linear module 101 and the linear module 101 are adjusted according to the different measurement tasks of the blade to be measured; The measurement task adjusts to different point cloud densities and performs scan detection in a specified area. In addition, according to the size and material reflective characteristics of different blades, it may be necessary to adjust the exposure and movement position for the first measurement of each blade to achieve the best measurement effect, for example, before the first measurement of the same blade batch begins , first move the two probes so that the laser light is on the surface of the blade. Adjust the appropriate exposure in the software so that the captured picture is suitable for the center extraction of the laser stripes. The movement position adjustment is to confirm whether the set measurement range can cover the entire blade. , and you can start measuring.
所述控制线性模组101和转台103运动，使视觉传感器102采集所述待测叶片的完整叶片点云数据，具体包括：控制线性模组101和转台103实现三轴联动，使视觉传感器102采集所述待测叶片的完整叶片点云数据。将待测叶片进行装夹在转台103上之后，两侧线性模组101上安装的视觉传感器102进行第一次向上扫描，随后转台103旋转指定角度，进行第一次向下扫描，按照以上扫描方式循环扫描直至采集到完整叶片点云。The control of the motion of the linear module 101 and the turntable 103, so that the visual sensor 102 collects the complete blade point cloud data of the blade to be tested, specifically includes: controlling the linear module 101 and the turntable 103 to realize three-axis linkage, so that the visual sensor 102 collects The complete blade point cloud data of the blade to be tested. After the blade to be tested is clamped on the turntable 103, the visual sensors 102 installed on the linear modules 101 on both sides perform the first upward scan, and then the turntable 103 rotates at a specified angle to perform the first downward scan. The scanning method is repeated until the complete blade point cloud is collected.
如图6所示，所述使视觉传感器102采集所述待测叶片的完整叶片点云数据，并对所述完整叶片点云数据进行处理，输出所述待测叶片的检测结果，具体包括：将两个视觉传感器102每次扫描的点云数据按照刚体变换的方法粗配准到同一个系统坐标系下，通过迭代最近点(Iterative Closest Point，ICP)方法对粗配准的点云进行精配准，并将精配准后的点云进行三维显示，用户可以详细查看扫描结果，然后，选择定制好的数字化样板，把所述精配准后的点云与相应的数字化样板进行比对，一键化输出可视化、定制化的叶片分析报告，其中，分析结果为由定制化的测量任务需求决定，自动输出含有表格、图片、自定义参数报告。其中，迭代最近点(Iterative Closest Point，ICP)方法具体为：根据配准点云和目标点云的初始位置，计算匹配对应点个数及对应点距离，使对应点距离及其他设置的阈值最小化，求出旋转移动矩阵，使配准点云移动，重复整个过程直至达到配准收敛要求。其中，ICP精配准无需每次手动设置阈值，默认情况下相应配准阈值将会根据每次的扫描点云进行自适应调整，在每次制作数字化样板时也可以直接给出最合适的阈值。As shown in FIG. 6 , the visual sensor 102 is used to collect the complete blade point cloud data of the blade to be tested, and the complete blade point cloud data is processed to output the detection result of the blade to be tested, which specifically includes: Roughly register the point cloud data scanned by the two visual sensors 102 into the same system coordinate system according to the method of rigid body transformation, and fine-tune the coarsely registered point cloud by the Iterative Closest Point (ICP) method. Registration, and three-dimensional display of the finely registered point cloud, the user can view the scanning results in detail, and then select a customized digital template to compare the finely registered point cloud with the corresponding digital template , One-click output of visualized and customized leaf analysis reports, in which the analysis results are determined by the requirements of customized measurement tasks, and automatically output reports containing tables, pictures, and custom parameters. Among them, the Iterative Closest Point (ICP) method is specifically: according to the initial position of the registration point cloud and the target point cloud, calculate the number of matching corresponding points and the distance between the corresponding points, so as to minimize the threshold of the corresponding point distance and other settings , find the rotation and movement matrix, and move the registration point cloud, and repeat the whole process until the registration convergence requirement is met. Among them, ICP fine registration does not need to manually set the threshold each time. By default, the corresponding registration threshold will be adaptively adjusted according to each scanning point cloud, and the most suitable threshold can also be directly given each time a digital template is made. .
如图7所示，本实施例还公开了一种应用上述叶片非接触式检测装置的叶片非接触式检测方法，该方法包括：As shown in FIG. 7 , this embodiment also discloses a non-contact blade detection method using the above-mentioned blade non-contact detection device, the method comprising:
S101、将待测叶片装夹在所述叶片非接触式检测装置的转台103上。S101. Clamp the blade to be tested on the turntable 103 of the blade non-contact detection device.
S102、控制柜104控制线性模组101和转台103实现三轴联动，使两个视觉传感器102进行第一次向上扫描，随后转台103旋转指定角度，两个视觉传感器102进行第一次向下扫描，按照以上扫描方式循环扫描直至采集到所述待测叶片的完整叶片点云数据。S102, the control cabinet 104 controls the linear module 101 and the turntable 103 to realize three-axis linkage, so that the two visual sensors 102 perform the first upward scan, and then the turntable 103 rotates at a specified angle, and the two visual sensors 102 perform the first downward scan , according to the above scanning method to scan circularly until the complete blade point cloud data of the blade to be tested is collected.
S103、控制柜104将两个视觉传感器102每次扫描的点云数据按照刚体变换的方法粗配准到同一个系统坐标系下，通过迭代最近点(Iterative Closest Point，ICP)方法对粗配准的点云进行精配准，并将精配准后的点云进行显示。S103, the control cabinet 104 roughly registers the point cloud data scanned by the two visual sensors 102 each time to the same system coordinate system according to the method of rigid body transformation, and uses the Iterative Closest Point (ICP) method to perform coarse registration fine registration of the point cloud, and display the fine registration point cloud.
S104、控制柜104把所述精配准后的点云与相应的数字化样板进行比对，输出可视化、定制化的叶片分析报告。S104. The control cabinet 104 compares the finely aligned point cloud with the corresponding digital template, and outputs a visualized and customized blade analysis report.
本发明提出的技术方案根据不同型号叶片的测量要求定制化测量分析结果，非接触式快速全面扫描叶片型面信息，自动处理采集的点云数据，与数字样板进行特征比较，输出相应叶片分析参数报告。通过将简易的操作与检测分析的自动化，大大提高叶片检测效率，改善大批量不同型号叶片检测的适应性。本发明具体优点如下：一、操作简单；操作人员只需将叶片装夹到工作平台105上的转台103上，在控制柜104的软件中进行叶片型号、测量项目等相关设置后，启动测量程序即可等待分析结果，简化检测流程降低了操作人员的操作难度；二、效率高；视觉传感器102根据非接触式激光三角测量原理，配合线性模组101以及转台103的运动，对视场内的待测叶片进行多角度全面扫描，由于视觉传感器102是由指定扫描位置触发扫描，整个扫描过程中分为向上向下两个方向，再一次提高了装置的整体扫描效率；三、自动化程度高；在测量过程，操作人员除了将叶片装夹到工作台上，只需要在软件中设置相应参数并启动测量即可，非接触式检测装置就会自动进行快速检测，并且输出相应检测分析报告，实现了检测分析一体化、自动化。The technical solution proposed by the invention customizes the measurement and analysis results according to the measurement requirements of different types of blades, quickly and comprehensively scans the blade profile information in a non-contact manner, automatically processes the collected point cloud data, compares the characteristics with the digital template, and outputs the corresponding blade analysis parameters Report. Through the automation of simple operation and detection analysis, the efficiency of blade detection is greatly improved, and the adaptability of large batches of different types of blade detection is improved. The specific advantages of the present invention are as follows: 1. The operation is simple; the operator only needs to clamp the blade on the turntable 103 on the working platform 105, and after performing related settings such as blade model and measurement items in the software of the control cabinet 104, start the measurement program You can wait for the analysis results, simplify the detection process and reduce the difficulty of the operator's operation; 2. High efficiency; the visual sensor 102 is based on the principle of non-contact laser triangulation, and cooperates with the movement of the linear module 101 and the turntable 103 to detect the objects in the field of view. The blade to be tested is fully scanned from multiple angles. Since the visual sensor 102 is triggered by the specified scanning position, the entire scanning process is divided into two directions, up and down, which once again improves the overall scanning efficiency of the device; 3. High degree of automation; During the measurement process, in addition to clamping the blades on the workbench, the operator only needs to set the corresponding parameters in the software and start the measurement. The non-contact detection device will automatically perform rapid detection and output the corresponding detection analysis report to realize Integration and automation of detection and analysis.
注意，上述仅为本发明的较佳实施例及所运用技术原理。本领域技术人员会理解，本发明不限于这里所述的特定实施例，对本领域技术人员来说能够进行各种明显的变化、重新调整和替代而不会脱离本发明的保护范围。因此，虽然通过以上实施例对本发明进行了较为详细的说明，但是本发明不仅仅限于以上实施例，在不脱离本发明构思的情况下，还可以包括更多其他等效实施例，而本发明的范围由所附的权利要求范围决定。Note that the above are only preferred embodiments of the present invention and applied technical principles. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and can also include more other equivalent embodiments without departing from the concept of the present invention, and the present invention The scope is determined by the scope of the appended claims.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810413725.1ACN108458659A (en) | 2018-05-03 | 2018-05-03 | A kind of blade contactless detection device and method |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810413725.1ACN108458659A (en) | 2018-05-03 | 2018-05-03 | A kind of blade contactless detection device and method |
| Publication Number | Publication Date |
|---|---|
| CN108458659Atrue CN108458659A (en) | 2018-08-28 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810413725.1APendingCN108458659A (en) | 2018-05-03 | 2018-05-03 | A kind of blade contactless detection device and method |
| Country | Link |
|---|---|
| CN (1) | CN108458659A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109489580A (en)* | 2018-12-10 | 2019-03-19 | 华东理工大学 | A kind of processing of complex surface in machine point cloud detection and compensation method |
| CN110864636A (en)* | 2019-11-21 | 2020-03-06 | 石家庄格力电器小家电有限公司 | Size detection device, size detection equipment and detection method thereof |
| CN110906874A (en)* | 2019-12-31 | 2020-03-24 | 华中科技大学无锡研究院 | Measurement method and system of end wall thickness based on steel pipe spiral travel |
| CN111515141A (en)* | 2020-04-28 | 2020-08-11 | 上海工程技术大学 | Automatic device for detecting part size and detection method |
| CN111982009A (en)* | 2020-02-26 | 2020-11-24 | 深圳市安达自动化软件有限公司 | Draw-bar box 3D size detection system and method |
| WO2020223594A3 (en)* | 2019-05-02 | 2020-12-17 | Kodak Alaris, Inc | Automated 360-degree dense point object inspection |
| CN112484660A (en)* | 2020-11-12 | 2021-03-12 | 中国航发南方工业有限公司 | Non-contact type engine blade profile detection method and engine blade |
| CN113405465A (en)* | 2021-06-17 | 2021-09-17 | 西北工业大学 | Engine blade tenon root width measuring device and measuring method |
| CN113791086A (en)* | 2021-09-08 | 2021-12-14 | 天津大学 | Method and device for surface defect measurement of sector blade based on computer vision |
| CN113838147A (en)* | 2021-09-29 | 2021-12-24 | 上海海事大学 | Blade assembly visual guide method and system based on depth camera |
| CN114543731A (en)* | 2020-11-25 | 2022-05-27 | 中国航发商用航空发动机有限责任公司 | Blade detection method and device |
| CN116086337A (en)* | 2023-04-06 | 2023-05-09 | 珞石（北京）科技有限公司 | Automatic propeller blade measuring device and method based on 3D vision |
| US12001191B2 (en) | 2017-11-17 | 2024-06-04 | Kodak Alaris Inc. | Automated 360-degree dense point object inspection |
| US12254617B2 (en) | 2017-11-17 | 2025-03-18 | Kodak Alaris Llc | Automated in-line object inspection |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8031346B2 (en)* | 2008-10-31 | 2011-10-04 | Siemens Energy, Inc. | Coating evaluation process |
| CN102288115A (en)* | 2011-07-11 | 2011-12-21 | 南京信息工程大学 | Method for precisely measuring roller geometric dimension on line |
| WO2015038354A1 (en)* | 2013-09-10 | 2015-03-19 | Faro Technologies, Inc. | Use of a three-dimensional imager's point cloud data to set the scale for photogrammetry |
| CN104515478A (en)* | 2014-12-11 | 2015-04-15 | 华中科技大学 | Automatic three-dimensional measuring method and automatic three-dimensional measuring system for high-precision blade of aviation engine |
| CN104567679A (en)* | 2015-01-08 | 2015-04-29 | 华中科技大学 | Turbine blade visual inspection system |
| CN204514271U (en)* | 2015-01-08 | 2015-07-29 | 华中科技大学 | A kind of system of turbo blade vision-based detection |
| CN105737762A (en)* | 2016-05-09 | 2016-07-06 | 中国民用航空飞行学院 | Aviation engine blade profile measuring method |
| CN106370106A (en)* | 2016-09-30 | 2017-02-01 | 上海航天精密机械研究所 | Industrial robot and linear guide rail-combined linear laser scanning measurement system and method |
| CN107246849A (en)* | 2017-05-25 | 2017-10-13 | 西安知象光电科技有限公司 | A kind of blade optics method for fast measuring based on the axle measuring system of double testing head four |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8031346B2 (en)* | 2008-10-31 | 2011-10-04 | Siemens Energy, Inc. | Coating evaluation process |
| CN102288115A (en)* | 2011-07-11 | 2011-12-21 | 南京信息工程大学 | Method for precisely measuring roller geometric dimension on line |
| WO2015038354A1 (en)* | 2013-09-10 | 2015-03-19 | Faro Technologies, Inc. | Use of a three-dimensional imager's point cloud data to set the scale for photogrammetry |
| CN104515478A (en)* | 2014-12-11 | 2015-04-15 | 华中科技大学 | Automatic three-dimensional measuring method and automatic three-dimensional measuring system for high-precision blade of aviation engine |
| CN104567679A (en)* | 2015-01-08 | 2015-04-29 | 华中科技大学 | Turbine blade visual inspection system |
| CN204514271U (en)* | 2015-01-08 | 2015-07-29 | 华中科技大学 | A kind of system of turbo blade vision-based detection |
| CN105737762A (en)* | 2016-05-09 | 2016-07-06 | 中国民用航空飞行学院 | Aviation engine blade profile measuring method |
| CN106370106A (en)* | 2016-09-30 | 2017-02-01 | 上海航天精密机械研究所 | Industrial robot and linear guide rail-combined linear laser scanning measurement system and method |
| CN107246849A (en)* | 2017-05-25 | 2017-10-13 | 西安知象光电科技有限公司 | A kind of blade optics method for fast measuring based on the axle measuring system of double testing head four |
| Title |
|---|
| 罗楠等: "《点云配准中初始变换的快速优化求解算法》", 《西安电子科技大学学报（自然科学版）》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12254617B2 (en) | 2017-11-17 | 2025-03-18 | Kodak Alaris Llc | Automated in-line object inspection |
| US12001191B2 (en) | 2017-11-17 | 2024-06-04 | Kodak Alaris Inc. | Automated 360-degree dense point object inspection |
| CN109489580B (en)* | 2018-12-10 | 2021-09-24 | 华东理工大学 | An on-machine point cloud detection and compensation method for surface machining of aero-engine blades |
| CN109489580A (en)* | 2018-12-10 | 2019-03-19 | 华东理工大学 | A kind of processing of complex surface in machine point cloud detection and compensation method |
| WO2020223594A3 (en)* | 2019-05-02 | 2020-12-17 | Kodak Alaris, Inc | Automated 360-degree dense point object inspection |
| CN110864636A (en)* | 2019-11-21 | 2020-03-06 | 石家庄格力电器小家电有限公司 | Size detection device, size detection equipment and detection method thereof |
| CN110906874A (en)* | 2019-12-31 | 2020-03-24 | 华中科技大学无锡研究院 | Measurement method and system of end wall thickness based on steel pipe spiral travel |
| CN111982009A (en)* | 2020-02-26 | 2020-11-24 | 深圳市安达自动化软件有限公司 | Draw-bar box 3D size detection system and method |
| CN111515141A (en)* | 2020-04-28 | 2020-08-11 | 上海工程技术大学 | Automatic device for detecting part size and detection method |
| CN112484660B (en)* | 2020-11-12 | 2022-08-26 | 中国航发南方工业有限公司 | Non-contact type engine blade profile detection method |
| CN112484660A (en)* | 2020-11-12 | 2021-03-12 | 中国航发南方工业有限公司 | Non-contact type engine blade profile detection method and engine blade |
| CN114543731A (en)* | 2020-11-25 | 2022-05-27 | 中国航发商用航空发动机有限责任公司 | Blade detection method and device |
| CN113405465A (en)* | 2021-06-17 | 2021-09-17 | 西北工业大学 | Engine blade tenon root width measuring device and measuring method |
| CN113791086A (en)* | 2021-09-08 | 2021-12-14 | 天津大学 | Method and device for surface defect measurement of sector blade based on computer vision |
| CN113838147A (en)* | 2021-09-29 | 2021-12-24 | 上海海事大学 | Blade assembly visual guide method and system based on depth camera |
| CN113838147B (en)* | 2021-09-29 | 2024-01-19 | 上海海事大学 | Blade assembly visual guiding method and system based on depth camera |
| CN116086337A (en)* | 2023-04-06 | 2023-05-09 | 珞石（北京）科技有限公司 | Automatic propeller blade measuring device and method based on 3D vision |
| Publication | Publication Date | Title |
|---|---|---|
| CN108458659A (en) | A kind of blade contactless detection device and method | |
| CN111750805B (en) | A three-dimensional measurement device and measurement method based on binocular camera imaging and structured light technology | |
| CN101182990B (en) | Machine vision based geometric measuring systems of large-sized workpieces in process | |
| CN108844459B (en) | Calibration method and device of blade digital sample plate detection system | |
| CA2781289C (en) | Projection aided feature measurement using uncalibrated camera | |
| CN112614098B (en) | Blank positioning and machining allowance analysis method based on augmented reality | |
| CN109612390A (en) | Large-size workpiece automatic measuring system based on machine vision | |
| CN102221331B (en) | Measuring method based on asymmetric binocular stereovision technology | |
| CN105806251A (en) | Four-axis measuring system based on line laser sensor and measuring method thereof | |
| CN109596073A (en) | A kind of origin position scaling method of the revolving platform central axis of view-based access control model measurement | |
| CN108827187B (en) | A 3D Profile Measurement System | |
| CN113532277B (en) | Method and system for detecting plate-shaped irregular curved surface workpiece | |
| CN103175485A (en) | Method for visually calibrating aircraft turbine engine blade repair robot | |
| CN106705847A (en) | Geometric size measuring method and device for slit inside free-form surface | |
| CN104551865A (en) | Image measuring system and method | |
| CN109272555B (en) | A method of obtaining and calibrating external parameters of RGB-D camera | |
| CN107121967A (en) | A kind of laser is in machine centering and inter process measurement apparatus | |
| JP7353757B2 (en) | Methods for measuring artifacts | |
| Zhu et al. | Noncontact 3-D coordinate measurement of cross-cutting feature points on the surface of a large-scale workpiece based on the machine vision method | |
| CN110657750B (en) | Detection system and method for passivation of cutting edge of cutter | |
| CN111702054A (en) | A kind of curved plate incremental forming in-situ shape adjustment detection system and method | |
| CN106767443A (en) | A kind of new automatic secondary element image detector and measuring method | |
| CN110375674A (en) | A visual inspection system for precision manufacturing equipment | |
| CN114777690A (en) | A kind of surface roughness automatic measuring device and method | |
| Katz et al. | Closed-loop machining cell for turbine blades |
| 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 | ||
| EE01 | Entry into force of recordation of patent licensing contract | Application publication date:20180828 Assignee:Jiangsu Jihui Huake Intelligent Equipment Technology Co., Ltd. Assignor:Wuxi research institute of the Central China University of Science and Technology Contract record no.:X2019980000288 Denomination of invention:Blade non-contact type detection device and method License type:Common License Record date:20191023 | |
| EE01 | Entry into force of recordation of patent licensing contract | ||
| RJ01 | Rejection of invention patent application after publication | Application publication date:20180828 | |
| RJ01 | Rejection of invention patent application after publication |