计算机与现代化

• 图像处理 • 上一篇    下一篇

基于动态阈值和修正模型的舌体提取算法

  

  1. 1.嘉应学院医学院,广东梅州514031;2.华南理工大学计算机科学与工程学院,广东广州510006
  • 出版日期:2014-11-27 发布日期:2014-12-10
  • 作者简介: 张志顺(1978-),男,广东梅县人,嘉应学院医学院讲师,硕士,研究方向:医学图像处理; 刘勇(1973-),男,华南理工大学计算机科学与工程学院副教授,博士,研究方向: 模式识别与图像处理。
  • 基金资助:
     广东省自然科学基金资助项目(9151064101000106); 梅州市科技计划项目(2012B73)

 Tongue Extraction Algorithm Based on Dynamic Thresholds and Correction Model

  1. 1. School of Medicine, Jiaying College, Meizhou 514031, China;

      2. School of Computer Science and Engineering, South China University of Technology, Guangzhou 510006, China
  • Online:2014-11-27 Published:2014-12-10

摘要:

针对在舌象中较难准确分离舌体的问题,提出从舌体的颜色、形状、位置等先验知识出发,采用HIS色彩模型去除舌象中的嘴唇、脸部等信息,用动态阈值分割算法提取舌体初始轮廓,用
舌体修正模型得到最终舌体。通过舌体提取实验和数据分析表明,该方法不仅在抗噪能力、提取精度方面好于其它算子,而且克服了嘴唇、脸部的影响,对舌体凹陷区域也取得了较好的分割效果。舌体
的准确提取为下一步舌质舌苔的准确分离打下了良好基础。

关键词:  , HIS色彩模型; 动态阈值; 修正模型; 图像融合; 舌体提取

Abstract:

 To address the problem of segmenting the tongue body from a tongue picture, a novel algorithm is proposed, which is based on prior knowledge of color, shape
and position of tongue body. Using HIS color model, the algorithm preserves the contour of tongue body while information of lip and face tissues is removed accurately. After
that, the tongue body contour is refined by dynamic threshold segmentation algorithm. Finally, the tongue body is extracted by a regularization model. Experiments and data
analysis show that the proposed method in the ability to resist noise and extract the precision better than other operator. In addition, the method works well in presence of lip
and face tissues. good results are obtained by the proposed algorithm in spite of the sunken area of tongue body. Accurate segmentation of tongue body plays an important role of
preparing good extraction of coating on the tongue.

Key words:  HIS color model, dynamic threshold, correction model, image fusion, tongue body extract