TISNet-Enhanced Fully Convolutional Network with Encoder-Decoder Structure for Tongue Image Segmentation in Traditional Chinese Medicine

<div>Difficult cases for automated tongue image segmentation. (a) Tongue bodies with different size and shape. (b) Severe fissures and teeth marks on the surface of a tongue body. (c) Similar color between tongue body and lips.</div>

Computational and Mathematical Methods in Medicine

fig2

Figure 2

Figure 2: TISNet-Enhanced Fully Convolutional Network with Encoder-Decoder Structure for Tongue Image Segmentation in Traditional Chinese Medicine 