痕量微生物图像的阈值分割方法研究  

Research on thresholding segmentation method for trace microbial distribution image

作　　者：毛佳文[1] 刘晓梅[2] 李抄[1] 陈锋[1] 韦婧[1] 吴太虎[1] 

机构地区：[1]军事医学科学院卫生装备研究所,天津300161 [2]武警后勤学院附属医院检验科,天津300161

出　　处：《光学技术》2016年第3期229-233,共5页Optical Technique

基　　金：国家十二五规划重大专项资助项目(AWS11Z005-4);国家科技重大专项资助项目(2012ZX10004801)

摘　　要：探讨阈值分割算法,把其应用于痕量微生物图像的处理。基于阈值算法提出了适合痕量微生物图像的目标信号提取方法。对阈值分割算法进行分析,对比传统阈值方法和提出的改进阈值算法的实际图像处理效果。传统阈值算法可实现对图像中微生物信号的提取,但存在一定失真。基于阈值分割方法的改进算法具有抑制噪声信号,及区分背景与目标的能力,可高效实现微生物信号的提取。所成微生物分布二值图像具有轮廓清晰、易于观察的特点。The threshold segmentation algorithm applied to deal with trace microorganisms image is discussed. The target signal from trace microbial image is extracted. The practical effect to deal with trace microorganism＇s image is compared and analyzed between traditional threshold algorithm and improved threshold algorithm. The image extracted form microbial which used traditional threshold algorithm exist distortion. Improved threshold segmentation algorithm can eliminate image of noise signal, efficient distinguish background and objectives signal, and extract microbial signal from image. The image have a clear outline and easy to observe.

关 键 词：微生物图像分割 阈值法 全局阈值 阈值选取 改进阈值算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]