TongueNet: A Precise and Fast Tongue Segmentation System Using U-Net with a Morphological Processing Layer

Jianhang Zhou,Xiaojiao Chen,Bob Zhang,Qi Zhang

doi:10.3390/app9153128

Abstract

Automated tongue segmentation is a critical component of tongue diagnosis, especially in Traditional Chinese Medicine (TCM), where it has been practiced for thousands of years and is generally considered pain-free and non-invasive. Therefore, a more precise, fast, and robust tongue segmentation system to automatically segment tongue images from its raw format is necessary. Previous algorithms segmented the tongue in different ways, where the results are either inaccurate or time-consuming. Furthermore, none of them developed a dedicated, automatic segmentation system. In this paper, we proposed TongueNet, which is a precise and fast automatic tongue segmentation system. U-net is utilized as the segmentation backbone applying a small-scale image dataset. Besides this, a morphological layer is proposed in the latter stages of the architecture. The proposed system when applied to a tongue image dataset with 1000 images, achieved the highest Pixel Accuracy of 98.45% and consumed 0.267 s per picture on average, which outperformed conventional state-of-the-art tongue segmentation methods in both accuracy and speed. Extensive qualitative and quantitative experiments showed the robustness of the proposed system concerning different positions, poses, and shapes. The results indicate a promising step in achieving a fully automated tongue diagnosis system.

Highlights

The human tongue is a large and soft piece of flesh found in the mouth and primarily used for tasting and speaking [1]
Taking the difficulties as well as the efficiency into consideration, we propose a morphological processing layer (MPL) to refine the coarse mask image produced by the network using designed filters
TongueNet is compared with Floodfill [28] and

Summary

Introduction

The human tongue is a large and soft piece of flesh found in the mouth and primarily used for tasting and speaking [1]. Besides its essential functions in the human digestive system, a tongue can act as a key region of interest in disease diagnosis using traditional medicines such as Traditional. Traditional Chinese Tongue Diagnosis (TCTD) [2,3,4,5,6,7] performs pain-free and non-invasive disease detection on our human bodies by analyzing the different attributes of the tongue (e.g., color, shape and texture). This has been practiced for thousands of years. To perform TCTD automatically, a computer-aided tongue diagnosis system was proposed [8] which contains tongue segmentation as one of the key procedures

Results

Discussion

Conclusion