摘要
针对传统的多方向灰度形态学边缘检测算法存在计算量大、效率低的缺点,提出了一种基于自适应噪声抑制的多方向灰度形态学图像边缘检测算法。根据图像所含噪声的种类,采用不同尺度的结构元素对图像进行分类滤波,再根据像素点间灰度值的变化确定边缘方向,由相应方向的结构元素进行边缘检测。实验结果表明,与传统的多方向灰度形态学边缘检测算法相比,检测到的边缘重构相似度和边缘置信度更高,边缘连续性更强,且计算量低,运行效率高。
Aiming at the disadvantages of large amount calculation and low efficiency of the traditional multi-directions gray-scale morphology edge detection algorithm, a multi-directions gray-scale morphology edge detection algorithm based on adaptive noise suppressing is proposed. According to the type of noise existing in image, the multi-scale struc- tural elements are applied for filtering to the image. The edge direction is determined according to the change between the pixel gray values. The image edge information is detected by the structural elements in the corresponding direction. Experimental results show that the edge reconstruction similarity, the edge confidence level and the edge continuity are better via this algorithm than the traditional multi-directions gray-scale morphology edge detection version, and the computation amount is decreased and the run efficiency is increased.
出处
《光学技术》
CAS
CSCD
北大核心
2016年第3期234-238,共5页
Optical Technique
基金
国家自然科学基金(61175090)
山东省自然科学基金(ZR2014EL024)
山东省高等学校科技计划项目(J14LN76)
关键词
噪声抑制
边缘检测
灰度形态学
结构元素
noise suppressing
edge detection
gray scale morphology
structural elements
