数学形态学在海洋浮游植物显微图像处理中的应用

Application of Mathematical Morphology to Microscope Image Processing of Marine Phytoplankton

数学形态学是分析几何形状和结构的数学方法,广泛应用于各类图像形状和结构的分析与处理.浮游植物是海洋生态系统中的重要初级生产者,其个体微小,但形态特征各异,一般要在显微镜下才能进行准确的种类鉴定.本文以海洋常见赤潮藻血红哈卡藻(Akashiwo sanguinea(Hirasaka)Hansen&Moestrup)为对象,利用数学形态学方法来处理和分析其细胞显微图像.数学形态学算法包括二值图像形态学和灰度图像形态学,涉及到膨胀、腐蚀、开闭运算,提供了快速高效的优化算法,特别是基于形态学分水岭的图像分割、标记目标、目标分割.研究结果证明,数学形态学可应用于海洋浮游植物显微图像处理,为浮游植物显微自动识别提供了理论基础.

Mathematical morphology is one of a mathematical methods for analysis of geometric shape and structure, and has been commonly used in image analysis and treatment. Phytoplankton is important primary producer in marine ecosystem whose cell is usually small and can be weU identified only under microscope although it is usually typical in morphology for different species. Akashiwo sanguinea （Hirasaka） Hansen ＆ Moestrup, a common red - tide causative phytoplankton species, is selected as tested target for image treatment with mathematical morphology in the present study. Different mathematical morphological methods were used in this study for analysis and processing of marine phytoplankton microscope images. The algorithms include binary image morphological processing algorithm and gray image morphological processing algorithm refen＇ing to quick and efficient optimized algorithms of erosion and dilation, opening - closing operation, especially the image segmentation base on morphological watershed,target tags and target segmentation. It is suggested that mathematical morphology could be a good method for microscope image treatment and is very likely to be applied in automatic identification system of marine phytoplanktons.