摘要
X射线管道焊缝探伤机器人作业时,要求管外X射线接收器跟随管内X射线发射窗同步旋转。针对此要求,文中研究了一种基于视觉的同步算法。实现同步旋转的难点在于如何实时、准确地检测管内外间的转角差,文中通过对X射线图像进行特征提取检测转角差:先对X射线图像进行形态学滤波,接着用自适应阀值算法对其二值化,然后进行边缘提取,最后通过Hough变换提取特征并获取转角差。文中对机器人作业时拍摄的一组X射线图像进行了试验,结果表明:文中算法从X射线图像中准确检测出了管内外间的转角差,且检测速度超过25帧/秒的实时性要求。文中算法再结合PID等规律控制伺服电机可实现管内外同步旋转。
When X - ray pipeline weld - inspection robot is working, the outer X - ray receiver is required to track the inner X - ray transmitter synchronously. In order to meet the requirement, this paper researches a kind of vision- based synchronous algorithms. The difficulty in realizing synchronization lies in how to detect the deviation angle between the outer receiver and the inner transmitter precisely and real - time, and this paper solves the problem by extracting features from X -ray images, as follows： firstly processes the image with a morphology filter, and then segments it with an adaptive threshold, and then processes it with edge detection, and finally extracts features form it with Hough transform and thus obtaining the deviation angle. Experiments on a series of X - ray images, which are taken by robot when it is working, are conducted. The experimental results show that, the algorithms detect the deviation angle precisely, and the detection velocity reaches 25 fps. Thus by controlling the servo - motor with the algorithms and a certain control rule such as PID, synchronization can be achieved.
出处
《计算机仿真》
CSCD
2006年第5期145-148,共4页
Computer Simulation
关键词
焊缝探伤
视觉同步
特征提取
形态学滤波
自适应阀值
Weld inspection
Visual synchronous
Feature extracting
Morphology filter
Adaptive threshold
