《计算机应用研究》|Application Research of Computers

基于图像区域特征的自适应边缘检测

Adaptive edge detection based on image region characteristics

免费全文下载 (已被下载 次)  
获取PDF全文
作者 刘付民,张治斌,王娜,沈记全
机构 河南理工大学 计算机科学与技术学院,河南 焦作 454000
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2012)06-2382-03
DOI 10.3969/j.issn.1001-3695.2012.06.102
摘要 边缘检测技术的关键在于检测前的图像去噪问题。Madenda方法考虑了尖锐、模糊、噪声的影响, 但在全图像范围内采用的是统一的噪声抑制参数和模糊控制参数, 从而缺乏对不同区域内噪声抑制的针对性。为此, 提出基于图像区域特征的自适应边缘检测方法, 通过熵和对比度的计算, 将原始图像逐级分解, 直到达到预先设定的阈值条件。然后根据各个区域的特点, 自适应地计算出相应的噪声抑制参数和模糊参数, 从而达到对该区域噪声最合理的抑制, 以达到最佳的边缘检测效果。实验结果表明, 该方法可以有效地抑制噪声的影响, 获得高质量的边缘检测图像。
关键词 边缘检测;区域分割;自适应;模糊参数
基金项目 河南省软科学研究计划资助项目(102400450034)
河南省科技计划基金资助项目(092300410216)
本文URL http://www.arocmag.com/article/01-2012-06-102.html
英文标题 Adaptive edge detection based on image region characteristics
作者英文名 LIU Fu-min, ZHANG Zhi-bin, WANG Na, SHEN Ji-quan
机构英文名 School of Computer Science & Technology, Henan Polytechnic University, Jiaozuo Henan 454000, China
英文摘要 The key problem is noise filtering before detection. Sharp, blur and noise are considered in Madenda method. But exclusive parameter to control noise and blur is adopted in the whole image. This parameter is hard to have efficiency to all kinds of noise in different region. This paper proposed adaptive edge detection based on image region characteristics, and segmented original image by calculating entropy and contrast until the threshold was satisfied. Then it computed the parameter to control noise and blur according to specialty in all regions. These parameters was best of all noise in different regions. Experiments results show that this proposed method proposed is effective to control noise and good edge detection image will be obtained.
英文关键词 edge detection; region segmentation; adaptive; blur parameter
参考文献 查看稿件参考文献
 
收稿日期
修回日期
页码 2382-2384
中图分类号 TP242
文献标志码 A