摘要
边缘检测是图像处理的重要环节之一,传统的边缘检测主要包含基于模板匹配和基于图像梯度两类方法。为了克服模板匹配方法边缘信息丢失较多和图像梯度方法易受图像噪声影响的缺点,该文给出了一种引导滤波自适应双阈值并改进Kirsch算子的优化边缘检测算法。针对图像局部信息特征,在图像的不同边缘位置动态生成引导滤波函数,从而保持和增强边缘的效果;同时对Kirsch算子复杂的运算量进行简化,根据图像边缘区域自适应地选择两个阈值,二者有效结合可大幅提高边缘检测算法的准确度和运算效率。实验结果表明,与传统边缘检测Kirsch算法和Sobel算法相比,该文算法的边缘定位和运算速度较优,能检测出精细的真实边缘,对图像的处理速度是传统算法的4倍以上。
The edge detection is one of the important aspects of image processing. The traditional edge detection method includes two kinds of methods,the method based on the template matching and the method based on the image gradient. In order to overcome the shortages that the template matching method losses more edge information and the image gradient method is easily affected by the noise,an adaptive double threshold modified Kirsch edge detection algorithm is proposed for the boot filtering. For the local information feature of the image,the boot filtering function is dynamicallygenerated at different edge positions of the image to maintain and enhance the edge effect. At the same time,the complex computation of the Kirsch operator is simplified,and the two thresholds are adaptively selected according to the image edge region threshold.The combination of the two thresholds can effectively improve the accuracy and the efficiency of the edge detection algorithm. The experimental results show that,compared with the traditional Kirsch algorithm and the Sobel algorithm,the improved algorithm is better in the edge localization and operation speed,and the processing speed of the image is more than 4 times of the traditional＇s while the fine real edge is well detected.
作者
许乐灵
胡石
Xu Leling;Hu Shi(Chizhou Vocational and Technical College,Chizhou 247000,Chin)
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2018年第2期177-182,共6页
Journal of Nanjing University of Science and Technology
基金
安徽省高等学校自然科学重点研究项目(KJ2018A0181)
安徽省高等学校质量工程重点研究项目(2016jyxm0715)
安徽省高校优秀青年人才支持计划项目(gxyq2017218)
关键词
引导滤波
自适应双阈值
边缘检测
boot filtering
adaptive double threshold
edge detection
