| 基于K邻近均值滤波器的自适应Canny边缘检测 |
| |
| 引用本文： | 刘建林,温显斌.基于K邻近均值滤波器的自适应Canny边缘检测[J].天津理工大学学报,2011,27(3):42-45. |
| |
| 作者姓名： | 刘建林 温显斌 |
| |
| 作者单位： | 天津理工大学计算机与通信工程学院天津市智能计算及软件新技术重点实验室，计算机视觉与系统教育部重点实验室，天津300384 |
| |
| 摘 要： | 本文在分析传统Canny算子的优缺点基础上,针对其在平滑滤波和阈值选取上的不足,提出了新的改进算法.该算法首先采用K邻近均值滤波器对图像进行预处理,然后利用非极大值抑制过程后图像本身的信息来自动产生高低阈值对图像进行边缘检测.经过实验证实,使用本算法检测到的边缘连续性更好,边缘细节更丰富,同时阈值选择的自适应能力有所提高.
|
| 关 键 词： | 图像处理 边缘检测 Canny算子 平滑滤波 非极大值抑制 |
New adaptive Canny edge detection based on K neighbor mean filter |
| |
| LIU Jian-lin,WEN Xian-bin.New adaptive Canny edge detection based on K neighbor mean filter[J].Journal of Tianjin University of Technology,2011,27(3):42-45. |
| |
| Authors: | LIU Jian-lin WEN Xian-bin |
| |
| Institution: | (School of Computer and Communications Engineering,Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology,Computer Vision and Systems Laboratory of the Ministry of Education,Tianjin University of Technology,Tianjin 300384,China) |
| |
| Abstract: | According to the traditional Canny operator,the advantages and disadvantages of the traditional Canny operator are analyzed,and in allusion to the lack of the selection of its filtering and threshold,new improved algorithm is proposed.The algorithm uses K neighbor mean filter to process the images,in the same time when noises are smoothed,the original gray features on the edge of the images are kept better;traditional threshold is artificial,lack of adaptability,and weak in automation capabilities.In this paper,we use the information of the image itself after non-maxima suppression process to generate high and low threshold automatically.Through experimental results,it is confirmed that it has better continuity of the edge detected by this algorithm,richer edge detail,and improved adaptive capacity. |
| |
| Keywords: | image processing edge detection Canny operator smoothing filter non-maxima suppression |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
