摘要
在丝带的生产过程中,要求对丝带的边缘进行检测以保证丝带的宽度一致。找出丝带中存在的卷边等现象,能有效地提高丝带生产的效率和质量。根据丝带图像自身的特点,分析出丝带背景对边缘的影响,在对图像进行灰度变换之后,用Otsu阈值分割法得到阈值,将该阈值作为Canny算子高门限进行丝带边缘检测。用灰度变换对丝带图像的像素范围进行线性变化,使得Otsu阈值分割法得到的阈值满足Canny算子实现边缘检测的要求。将该算法运用在丝带边缘检测系统中,达到了很好的检测效果,并且由于该改进的算法运算简单,检测速度能满足系统的要求,也实现了丝带检测系统的智能化。
In the production process of ribbon, it is required to detect the edge of ribbon so as to ensure consistent ribbon width. Finding hemming etc, in ribbon can effectively improve the production efficiency and quality of ribbon. This paper analyzes the influence of ribbon background on the edge according to features of ribbon image, obtains threshold value with Otsu threshold segmentation method after grey level transformation of image and detects ribbon edge with the threshold value as the high threshold of Canny operator; uses grey level transformation for linear variation of the pixel range of ribbon image and makes the threshold value obtained with Otsu threshold segmentation method meet the requirement of Canny operator for achieving edge detection; applies this algorithm in in ribbon edge detection system and achieves a good detection effect. Due to simple operation of the improved algorithm and its detection speed can meet the requirement of the system, intellectualization of ribbon test system has also been achieved.
出处
《丝绸》
CAS
北大核心
2013年第9期36-40,共5页
Journal of Silk
基金
南通市专项基金项目(2009-03)
关键词
图像分割
OTSU算法
CANNY算子
灰度变换
丝带检测
image segmentation
Otsu algorithm
Canny operator
grey level transformation
ribbon detection
