基于辅助光源的舌象点刺识别方法研究  

A Research about Tongue-Prickled Recognition Method Based on Auxiliary Light Source

作　　者：王学民[1,2] 王瑞云[1] 郭丹[1] 陆小左[3] 周鹏[1,2] WANG Xueming WANG Ruiyun GUO Dan LU Xiaozuo ZHO U Peng(School of Precision Instruments and Optical Electronics Engineering, Tianjin University, Tianfin 300072, China Tianjin Key Laboratory of Biomedical Detecting Techniques ＆ Instruments, Tianjin 300072, China College of Traditional Chinese Medicine, Tianjin University of Traditional Chinese Medicine, Tianjin 300193, China)

机构地区：[1]天津大学精密仪器与光电子工程学院,天津300072 [2]天津市生物医学检测技术与仪器重点实验室,天津300072 [3]天津中医药大学中医工程学院,天津300193

出　　处：《传感技术学报》2016年第10期1553-1559,共7页Chinese Journal of Sensors and Actuators

基　　金：国家"十二五"支撑计划项目(2012BAI25B05)

摘　　要：中医舌诊中点刺所生部位和其在不同部位的疏密程度对疾病诊断有重要价值,点刺自动识别是实现舌诊客观化诊断的关键和难点之一。针对现有点刺识别方法适应性不高、识别率低等问题,提出了一种基于辅助光源的点刺识别方法。对标准白色光源和纯绿色光源下采集的10例舌象分别进行分割、配准,在绿光舌象上实现点刺提取并显示在白光舌象上,算法自动提取正确率能达到专家识别率88.47%;实现点刺的自动区域定位和计数。实验结果表明该方法点刺识别率较高,可在临床领域有所应用,为多种疾病提供了重要的诊断信息。Prickle＇s growing area and density in different areas have important disease diagnosis value in Tradition-al Chinese Medicine＇s tongue inspection. Prickle automatic recognition is the key and difficulty to achieve tongueinspection objective diagnosis. Contrapose the problem existing prickled recognition methods have low recognitionrate and narrow adaptability,put forward a novel method to recognize prickle based on auxiliary light source. 10 tongue images were collected under standard white light source and pure green light source. Tongue bodies weresegmented and registered in white images and green images. Prickles were extracted in green images and showed inwhite images. In contrast to expert recognition rate,algorithm automatic extraction accuracy could reach about88.47%. And prickles could be automatically located and counted in different areas. The processing result showsthe presented method can attain high prickled recognition rate,realize accurately automatic quantitative analysisand then provide important diagnostic information for multiple diseases.

关 键 词：中医舌诊 点刺识别 图像分割 图像配准 纯绿色LED 

分 类 号：TP391.5[自动化与计算机技术—计算机应用技术]