基于OpenCV的输送带跑偏自动检测算法研究

Research of automatic detection algorithm of conveying belt deviation based on OpenCV

为提高输送带跑偏检测方法的实时性和精度,结合OpenCV函数库的优点,提出了一种输送带跑偏自动检测算法。该算法利用CCD摄像机实时采集输送带视频流图像,并对图像进行预处理,然后使用改进的Canny边缘检测算法检测图像边缘,最后利用Hough直线变换提取输送带跑偏特征,通过输送带图像的几何特征来判断其是否跑偏,如跑偏则报警。实验结果表明,该算法简单有效,运算速度快,实现了输送带跑偏的自动检测。

In order to improve real-time performance and precision of detection method of conveying belt deviation, an automatic detection algorithm of conveying belt deviation was proposed combined with the advantages of OpenCV. Firstly, the algorithm uses CCD camera to capture video stream image of conveying belt real timely, and pre-processes the video stream image. Then it uses improved Canny edge detection algorithm to detect image edge. Finally, it adopts Hough linear transform to extract the belt deviation characteristics, and determine whether the belt is deviation or not, and make alarming if deviation. The test results show that the algorithm is simple and effective, has high speed of operation, and realizes automatic detection of conveying belt deviation.