基于二维最大类间方差与数学形态学的舌像分割  

Image Segmentation of Tongue Image Based on Two-dimensional OTSU Algorithm and Mathematical Morphology

作　　者：高清河[1,2] 刚晶[1] 王和禹[1] 刘海英[1] GAO Qinghe;GANG Jing;WANG Heyu;LIU Haiying(College of Information Engineering,Liaoning University of Traditional Chinese Medicine,Shenyang 110847;College of Science,Northeastern University,Shenyang 110004)

机构地区：[1]辽宁中医药大学信息工程学院,沈阳110847 [2]东北大学理学院,沈阳110004

出　　处：《计算机与数字工程》2017年第6期1200-1203,共4页Computer & Digital Engineering

基　　金：辽宁省教育厅科学研究一般项目(编号:L2015333)资助

摘　　要：舌诊是中医诊断疾病的重要方法,实现中医舌诊的客观化和定量化是关系中医舌诊继承和发扬的重要工作。为了将舌体区域正确地从采集的舌像中分割出来,论文采用了OTSU算法与数学形态学相结合的方法对舌像进行分割,最大类间方差法用于对原图进行二值化增强处理,而数学形态学方法用于修正二值图像的边界,最终提取出舌体边缘。实验表明,经过OTSU算法和数学形态学进行图像分割,最后得到的舌体图像更加清晰、连贯,提高了图像的可视性和准确性。Tongue diagnosis is one of the essential methods of traditional Chinese medicine diagnosis.The realization of char?acterization and quantification of cdiagnosis is an important work which is related of succession and development of traditional Chi?nese tongue diagnosis.In order to obtain the tongue body area from the tongue image,the OTSU algorithm and mathematical mor?phology are introduced to segment tongue image.The OTSU algorithm is used to operate binary enhancement processing for the origi?nal image.However,the mathematical morphology is introduced to correct the boundary for binary image.Experimental results showthat the proposed method can get better segmentation effect for tongue image.

关 键 词：最大类间方差 数学形态学 图像滤波 舌像分割 

分 类 号：R318[医药卫生—生物医学工程] TP391[医药卫生—基础医学]