融合多颜色分量的舌图像阈值分割算法研究
|
摘要
针对中医自动化舌诊中的舌图像分割问题,提出一种融合多颜色分量的舌图像阈值分割算法。对RGB颜色空间中的蓝色和红色分量执行阈值分割,确定舌图像中的人脸区域;对HSI颜色空间中的色调分量执行变换,在变换后的色调分量上执行阈值分割,以获得包含真实舌体与上嘴唇的初始目标区域;对初始目标区域对应的红色通道执行阈值分割,得到舌根和嘴唇之间的间隙区域;利用间隙区域剔除掉初始目标区域中的上嘴唇,获得最终舌体分割结果。仿真实验表明:该算法较大程度地改善了舌图像分割的精度。
To solve the problem of tongue image segmentation in TCM automatic tongue diagnosis, we proposed a threshold segmentation algorithm of tongue image based on multi-color components. Threshold segmentation of blue and red components in RGB color space was performed to determine the face area in tongue image. The hue component in HSI color space was transformed, and the threshold segmentation was performed on the transformed hue component to obtain the initial target area including the real tongue body and upper lip. Threshold segmentation was performed on the red channel corresponding to the initial target area to obtain the gap between the tongue root and the lip. The upper lip in the initial target area was removed by using the gap area, and the final result of tongue segmentation was obtained. The simulation results show that the proposed algorithm greatly improves the accuracy of tongue image segmentation.
To solve the problem of tongue image segmentation in TCM automatic tongue diagnosis, we proposed a threshold segmentation algorithm of tongue image based on multi-color components. Threshold segmentation of blue and red components in RGB color space was performed to determine the face area in tongue image. The hue component in HSI color space was transformed, and the threshold segmentation was performed on the transformed hue component to obtain the initial target area including the real tongue body and upper lip. Threshold segmentation was performed on the red channel corresponding to the initial target area to obtain the gap between the tongue root and the lip. The upper lip in the initial target area was removed by using the gap area, and the final result of tongue segmentation was obtained. The simulation results show that the proposed algorithm greatly improves the accuracy of tongue image segmentation.
引文
[1] Kirschbaum B.Altas of Chinese tongue diagnosis[M].Seattle,WA:Eastland,2000.
[2] Yukii Junko.Tongue Diagnosis:Health password on the tip of the tongue[M].Beijing:Chemical Industry Press,2014.
[3] Hsu C H,Yu M C,Lee C H,et al.High eosinophil cationic protein levelin asthmatic patients with Heat Zheng[J].American Journal of Chinese Medicine,2003,31(2):277-283.
[4] Anastasi J K,Currie L M,Kim G H.Understanding diagnostic reasoning in TCM practice:tongue diagnosis[J].Alternative Therapies in Health & Medicine,2009,15(3):18-28.
[5] Abe S,Ishihara K,Adachi M,et al.Tongue-coating as risk indicator for aspiration pneumonia in edentate elderly[J].Archives of Gerontology and Geriatrics,2008,47(2):267-275.
[6] Hoefert S,Schilling E,Philippou S,et al.Amyloidosis of the tongue as a possible diagnostic manifestation of plasmacytoma[J].Mund Kiefer Gesichtschir,1999,3(1):46-49.
[7] Pang B,Zhang D,Li N,et al.Computerized tongue diagnosis based on Bayesian networks[J].IEEE Transactions on Biomedical Engineering,2004,51(10):1803-1810.
[8] Zhang H,Zuo W,Wang K,et al.A snake‐based approach to automated segmentation of tongue image using polar edge detector[J].International Journal of Imaging Systems & Technology,2010,16(4):103-112.
[9] Ning J,Zhang D,Wu C,et al.Automatic tongue image segmentation based on gradient vector flow and region merging[J].Neural Computing & Applications,2012,21(8):1819-1826.
[10] Shi M J,Li G Z,Li F F.C2G2FSnake:automatic tongue image segmentation utilizing prior knowledge[J].Science China Information Sciences,2013,56(9):1-14.
[11] Shi M J,Li G Z,Li F F,et al.Computerized tongue image segmentation via the double geo-vector flow[J].Chinese Medicine,2014,9(1):7-16.
[12] Otsu N.A Threshold Selection Method from Gray-Level Histograms[J].IEEE Transactions on Systems,Man,and Cybernetics,2007,9(1):62-66.
[13] Yasnoff W A,Mui J K,Bacus J W.Error measures for scene segmentation[J].Pattern Recognition,1977,9(4):217-231.
[2] Yukii Junko.Tongue Diagnosis:Health password on the tip of the tongue[M].Beijing:Chemical Industry Press,2014.
[3] Hsu C H,Yu M C,Lee C H,et al.High eosinophil cationic protein levelin asthmatic patients with Heat Zheng[J].American Journal of Chinese Medicine,2003,31(2):277-283.
[4] Anastasi J K,Currie L M,Kim G H.Understanding diagnostic reasoning in TCM practice:tongue diagnosis[J].Alternative Therapies in Health & Medicine,2009,15(3):18-28.
[5] Abe S,Ishihara K,Adachi M,et al.Tongue-coating as risk indicator for aspiration pneumonia in edentate elderly[J].Archives of Gerontology and Geriatrics,2008,47(2):267-275.
[6] Hoefert S,Schilling E,Philippou S,et al.Amyloidosis of the tongue as a possible diagnostic manifestation of plasmacytoma[J].Mund Kiefer Gesichtschir,1999,3(1):46-49.
[7] Pang B,Zhang D,Li N,et al.Computerized tongue diagnosis based on Bayesian networks[J].IEEE Transactions on Biomedical Engineering,2004,51(10):1803-1810.
[8] Zhang H,Zuo W,Wang K,et al.A snake‐based approach to automated segmentation of tongue image using polar edge detector[J].International Journal of Imaging Systems & Technology,2010,16(4):103-112.
[9] Ning J,Zhang D,Wu C,et al.Automatic tongue image segmentation based on gradient vector flow and region merging[J].Neural Computing & Applications,2012,21(8):1819-1826.
[10] Shi M J,Li G Z,Li F F.C2G2FSnake:automatic tongue image segmentation utilizing prior knowledge[J].Science China Information Sciences,2013,56(9):1-14.
[11] Shi M J,Li G Z,Li F F,et al.Computerized tongue image segmentation via the double geo-vector flow[J].Chinese Medicine,2014,9(1):7-16.
[12] Otsu N.A Threshold Selection Method from Gray-Level Histograms[J].IEEE Transactions on Systems,Man,and Cybernetics,2007,9(1):62-66.
[13] Yasnoff W A,Mui J K,Bacus J W.Error measures for scene segmentation[J].Pattern Recognition,1977,9(4):217-231.