摘要
针对藻细胞原始图像的特点,将数学形态学中的闭运算、自适应形态面积开运算、形态重构及Canny算子相结合,用于对混杂大量噪声的原始藻细胞灰度图像进行预处理,包括去噪、重建保留原始图像细节信息的藻细胞图像。在预处理的过程中,利用形态边缘检测算子对藻细胞的边缘轮廓进行了检测。实验证明该方法能有效的去噪。所提取的藻细胞的边缘轮廓精确、完整,为后续的特征提取及分类识别奠定了良好基础。
In this paper, the closing operation of mathematical morphology, adaptive area-open operation and morphological reconstruction in combination with Canny operator are used to preprocess the original grey-level image of an algal cell with noise. The preprocessing accomplishes the task of noise elimination and reconstruction of algal cell according to the characteristics of the original image of the algal cell. In the procedure of preprocessing, the morphology edge detection technique is applied. Experimental results demonstrate that the preprocessing method proposed in this paper is efficient in noise elimination while retaining lots of detailed information of the original image of the algal cell. The edge detected represents the outline of the cell accurately.
出处
《中国海洋大学学报（自然科学版）》
CAS
CSCD
北大核心
2008年第3期513-516,502,共5页
Periodical of Ocean University of China
基金
国家自然科学基金项目(60572064)资助
关键词
数学形态学
面积开
形态重构
mathematical morphology
area-open
morphological reconstruction
