摘要视觉测量系统是电力机器人自主作业的关键。本文通过对电力机器人工作过程的分析，对其视觉测量系统进行了总体方案设计。 应用张氏标定法进行了相机的标定，并利用标定得到的内外参数，针对左右图像极线没有对准的问题进行了图像立体校正。针对电力机器人在作业时采集的图像易受光照和环境中其他噪声影响的问题，采用了一种自适应阈值选取的Canny边缘检测方法对获得的图像进行边缘提取。以电力机器人经常遇到的作业对象螺栓为目标，分别使用传统的基于最小二乘法椭圆拟合与改进的椭圆拟合方法对其进行识别，比较并分析了两种方法的识别结果。使用不同的基于灰度的模板匹配方法进行了模板匹配实验，然后利用基于双目视觉的三维测量原理对目标空间位置进行了测量。实验结果表明所采用的方法能准确地测量出作业目标的空间位置。78571

毕业论文关键词  相机标定  Canny  椭圆拟合  模板匹配  视觉测量

毕业设计说明书外文摘要

Title   On Visual Measurement Methods of Power Robot   

Abstract Visual measurement system is the key of autonomous operation of power robot。 In this paper, through the analysis of the working process of the power robot, the overall scheme of the vision measurement system is designed。 The camera calibration is implemented using the Zhang's calibration method, and the image stereo calibration have been done for the left and right images polar alignment problem using the internal and external parameters which were obtained from the camera calibration。 Aiming at the problem of the image obtained by the power robot in the operation, which is subject to the influence of the illumination and the other noise in the environment, an adaptive threshold selection method of Canny edge detection is adopted to obtain the edge of the image。 Bolt which is often encountered in the work of power robot is used as a target for recognition。 The traditional ellipse fitting method based on least squares ellipse fitting and improvement on the recognition have been done for recognizing the target, respectively, and then the recognition results of two methods have been compared and analyzed。 Template matching experiments were carried out using different gray level based template matching method, and then the 3D measurement principle based on binocular vision was used to measure the space position of the object。 The experimental results showed that the method used can accurately measure the spatial position of the target。

Keywords  Camera calibration   Canny   Ellipse fitting   Template match Visual measurement

目   次

1  绪论 1

1。1  研究背景及意义 1

1。2  电力机器人国内外研究现状 1

1。2。1  国外研究现状 1

1。2。2  国内研究现状 2

1。3  视觉测量关键技术 2

1。3。1  相机标定 2

1。3。2  目标识别 3

1。4  本文内容安排 3

2  视觉测量系统总体设计 5

2。1  视觉测量系统需求分析 5

2。2  视觉测量系统组成 6

2。2。1  目标识别与测量系统方案设计 6

2。2。2  基于双目视觉的目标识别与测量实验系统设计