摘要
针对管道相贯线焊缝难以检测的问题,提出了一种用于管道相贯线焊缝自动检测机器人轨道识别和定位的图像处理算法。在颜色空间转换和彩色边缘检测基础上,引入图像金字塔方法的Hough变换直线检测算法,快速识别磁性色码条并实现机器人自身姿态和运动轨迹的实时控制,将多阈值处理引入区域生长算法分割人工标识点,对人工标识点进行计数从而实现定位。实验结果验证了提出的图像处理算法的有效性。
An image processing algorithm for automatic detecting robot of intersecting line weld to identify pathway and localization is proposed, aiming at problem that the intersecting line weld is difficult to be detected. On the basis of color space conversion and color edge detection, an image pyramids method is introduced to the Hough transform line detection algorithm ,to identify the magnetic color-coded strips quickly and achieve real-time control of the attitude and motion trajectory of the robot; a multi-threshold calculation is applied to the region growing algorithm to segment the artificial mark points, and the artificial mark points are counted so as to realize localization. The effectiveness of the proposed image processing algorithm is verified through experimental results.
出处
《传感器与微系统》
CSCD
2017年第7期146-149,153,共5页
Transducer and Microsystem Technologies
基金
深圳市技术创新计划资助项目(20150327135346)
关键词
图像处理
区域生长
多阈值
相贯线焊缝
焊缝检测机器人
image processing
region growing
multi-threshold
intersecting line weld
weld inspection robot
