摘要
提出了一种高效的在实时视频图像中检测轧钢钢带边缘的方法,该方法采用局部图像处理策略以降低计算量,根据钢带图像特征,设计了一种空间滤波器,在消除噪声的同时,能够保留较多边缘信息。同时基于计算出的边缘点位置的二维分布特征提出了一种简单、有效的边缘位置判断算法,从而可实时测出钢带的位置偏移量。并构造了实际的图像采集、处理和控制系统,取得良好的效果。
This paper proposes a high efficient method for detecting edge of steel blot form video image.Local image processing method is adopt to reduce amount of calculation,a kind of spatial filtering palate is designed based on characteristic of the image of steel bolt which could be able to keep more information but much less noise.An simple but efficient algorithm for judging position of the edge is designed based on 2-dimentional distribution characteristic,so that offset of position of steel blot could be calculated real timely.A practical system for image acquisition,image processing and online control is developed.Effect of the system is proved to be very good by practical test.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第20期219-221,共3页
Computer Engineering and Applications
