自适应参数优化的全自动舌像提取方法  

Full-Automatic Tongue Image Extraction Method with Adaptive Parameter Optimization

作　　者：朱明峰[1] 杜建强[1] 何雁[1] 张康[1] 何扬名[1] 朱懿敏[1] ZHU Mingfeng;DU Jianqiang;HE Yan;ZHANG Kang;HE Yangming;ZHU Yimin(School of Computer,Jiangxi University of Chinese Medicine,Nanchang 330004,China)

机构地区：[1]江西中医药大学计算机学院,南昌330004

出　　处：《江西中医药大学学报》2024年第6期10-12,16,共4页Journal of Jiangxi University of Chinese Medicine

基　　金：江西省中医药管理局科技计划项目(2022B988)。

摘　　要：目的:针对已有舌像提取方法存在参数多,每次需要手动设置的问题,设计一种自适应参数优化的全自动舌像提取方法。方法:分别采用区域搜索最优种子算法、色调直方图分析算法和亮度模式识别算法实现自适应参数优化,从而使舌像提取方法完全自动化,可有效改善用户体验,避免手动设定参数的繁琐和不确定性。结果:实验结果证实,本方法的舌像提取结果较为准确,对多种类型的舌像均能准确提取。结论:改进的方法可为舌诊客观化的后续检测方法提供科学有效的依据。Objective:Aiming at the problem that there are several parameters to be set by manual in the current tongue image extraction method,this paper has realized a kind of full automatic tongue image extraction method with adaptive parameter optimization.Methods:We separately adopt the algorithm of region search of optimal seed,the algorithm of hue histogram analysis and the algorithm of intensity pattern recognition,so as to realize the full automation of the tongue image extraction method,effectively improve the experience of users and avoid the complexity and uncertainty of manually setting parameters.Results:As the experiment results show,our method is relatively accurate,and can accurately extract multiple kinds of tongue images.Conclusion:The improved method can be considered as the basis of the following checking methods of objective tongue diagnosis.

关 键 词：自适应参数优化 D搜索算法 舌像提取 

分 类 号：R241.25[医药卫生—中医诊断学] TP301.6[医药卫生—中医临床基础]