一种有效的舌体分割算法研究

Study on an Effective Tongue Body Segmentation Algorithm

针对图像的色调值容易被多种因素影响,导致基于图像色调的舌体分割算法无法准确分割舌体,提出了一种基于新型颜色通道的自适应舌象分割方法。首先结合HSI颜色空间的S通道与I通道生成两个新的颜色通道,并将两个新的颜色通道二值化。然后提出一种自适应图像拼接方法,将两个二值图像进行拼接。接着将图片中非舌体区域的连通域排除,再将舌体区域进行形态学修正得到最终舌体分割结果。实验结果表明该方法具有较好的分割结果。

An adaptive tongue segmentation method Based on new color channels is proposed to solve the problem that the hue value of the image is easily affected by a variety of factors, resulting in the tongue segmentation algorithm that Based on hue value can not segment the tongue accurately.First, the S channel and the I channel of the HSI color space are combined to generate two new color channels and binarize the two new color channels.Secondly, an adaptive image mosaic method is proposed to splice two binary images. Thirdly, the connected domain of the non-tongue region in the picture is excluded, and the resulting tongue region is morphologically corrected to obtain the final tongue segmentation result. The experimental results show that the method has better segmentation results.