摘要
针对经典边缘检测方法中存在的精度低、算法复杂等问题,提出了一种基于维纳滤波的图像边缘检测方法。该方法不仅算法简单,而且可以较准确地检测出图像的灰度变化信息,不丢失重要边缘且实际边缘与检测到的边缘之间偏差较小。另外,利用该方法得到的边缘连续性较好,不易产生线段断裂。实验证明该方法可以用于检测图像边缘,尤其对以字符、曲线为主的复杂图像,有较好的检测效果。
Classical methods for edge detection have many disadvantages, such as low precision, complicated arithmetic etc. An approach for edge detection of images based on Wiener Filter is proposed, which is a simple one and the sensitivity to the variety of gray is excellent. Owing to these features, the obtained deviation between the real edge and measured one is a relatively small value without losing crucial edge. Moreover, the continuous edge can be obtained which has no breakpoint on line. The experiment results indicate that this approach is feasible, especially for those images basically composed with characters and curves.
出处
《工程图学学报》
CSCD
北大核心
2008年第6期91-95,共5页
Journal of Engineering Graphics
关键词
计算机应用
边缘检测
维纳滤波
图像技术
computer application
edge detection
Wiener filter
image technology
