舌图像自动分割方法

An Automatic Approach for Tongue Image Segmentation

提出了一种基于亮度信息和形态特征的舌图像自动分割方法。该方法首先利用大津法在亮度范围内自动选取阈值，进行二值化，然后通过对图像的像素进行标记，剔除非目标区域，最后利用数学形态学的方法获得最终的分割结果。试验结果表明，本方法具有令人满意的分割性能。

This paper presents an automatic approach for tongue image segmentation, which is based on the luminance information and the morphology features of the images. First, thresholding is done to the tongue image according the value determined by Otsu method in luminance field. Then, no-object regions are eliminated through labeling the pixels in the images. Finally, the ultimate result is produced with mathematical morphology. The experimental results show that this approach is effective in segmenting tongue image.