基于OpenCV的机器视觉尺寸测量研究  

Research on measurement of machine vision based on OpenCV

作　　者：郭瑞峰[1] 袁超峰[1] 杨柳[1] 彭光宇[1] GUO Ruifeng;YUAN Chaofeng;YANG Liu;PENG Guangyu(School of Mechanical and Electrical Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, China)

机构地区：[1]西安建筑科技大学机电工程学院,西安710055

出　　处：《计算机工程与应用》2017年第9期253-257,共5页Computer Engineering and Applications

基　　金：陕西省工业科技攻关项目(No.2015GY068)

摘　　要：通过对激光测距传感器与机器视觉的研究利用激光测距传感器、视觉传感器相结合的方法,提出了一种不同于结构光和双目尺寸测量的更为简单有效的新方法和新技术。首先对灰度直方图比较分析选取了最佳滤波方法,然后通过激光测距传感器测距和单目视觉传感器进行边缘的精确检测。激光测距传感器实现了对采样目标车厢进行快速测距定位,计算出采样机械手与目标车辆的相对位置关系。视觉传感器准确识别车厢尺寸及定位,为煤炭采样机械手自动采样提供了实现基础。检测目标是拉煤货运火车车厢,给煤炭采样行业带来更为简单有效,准确的目标定位的识别系统和方法。Through the study of laser ranging sensor and machine vision using laser ranging sensor,vision sensor,based on a combination of methods,this paper proposes a different in structured light and binocular size measurement of more simple and effective new method and new technology.Firstly,the optimal filtering method is selected by comparing the gray histogram,then the edge detection is performed by the laser ranging sensor and the monocular vision sensor.The laser ranging sensor realizes the fast ranging and positioning of the sampled targets,and calculates the relative position between the sampling manipulator and the target vehicle.The vision sensor can accurately identify the size and location of the car,which provides the basis for the automatic sampling of coal sampling manipulator.Detection goal is pull coal freight train,for the coal sampling industry to bring more simple,effective and accurate target location identification system and methods.

关 键 词：机器视觉 OPENCV 边缘检测 尺寸测量 

分 类 号：TP246.62[自动化与计算机技术—检测技术与自动化装置]