Fast marching over the 2D Gabor magnitude domain for tongue body segmentation

Abstract

Tongue body segmentation is a prerequisite to tongue image analysis and has recently received considerable attention. The existing tongue body segmentation methods usually involve two key steps: edge detection and active contour model (ACM)-based segmentation. However, conventional edge detectors cannot faithfully detect the contour of the tongue body, and the initialization of ACM suffers from the edge discontinuity problem. To address these issues, we proposed a novel tongue body segmentation method, GaborFM, which initializes ACM by performing fast marching over the two-dimensional (2D) Gabor magnitude domain of the tongue images. For the enhancement of the contour of the tongue body, we used the 2D Gabor magnitude-based detector. To cope with the edge discontinuity problem, the fast marching method was utilized to connect the discontinuous contour segments, resulting in a closed and continuous tongue body contour for subsequent ACM-based segmentation. Qualitative and quantitative results showed that GaborFM is superior to the other methods for tongue body segmentation.

Research and application of tongue and face diagnosis based on deep learning

“…Shi et al 109 proposed a double geo-vector flow (DGF) based tongue edge detection method, which is able to detect tongue edges and segment tongue regions for geological gradient vector flow evaluation of the tongue. Cui et al 110 proposed a new tongue segmentation method, GaborFM. This method uses a fast-marching algorithm to connect discontinuous contour segments to form a closed continuous tongue contour for ACM-based tongue segmentation.…”

Interactive Tongue Body Segmentation

“…By the GaborFM, we obtained the mask of tongue body, reliable lines and the initial tongue body segmentation. From the experiment section in [13,14], we can achieve the contours for the most tongue body images by using GaborFM. as shown in Figure 1, one can see that GaborFM is effective for the general tongue images, but it cannot detect the contour of tongue body, which is so weak, as shown in Figure 1(d).…”

“…Following the three guidelines, we proposed the interactive method. First we achieved a tongue body segmentation using the automatic segmentation method: GaborFM [13,14]. Then if the results cannot meet the requirement of the user, the method would update the segmentation materials, including tongue body mask, reliable-lines, using the contour information provided by the user.…”

“…We propose an interactive method by extending the GaborFM [13,14] for the tongue body segmentation, which can solve the problem of tongue body segmentation completely. Considering the limitations of existing interactive image segmentation methods, we proposed three guidelines: convenience, convergence property and high efficiency.…”

“…An automatic tongue body segmentation method we proposed before, GaborFM [13,14], is used for initial tongue body segmentation. If the initial tongue body segmentation can meet the user's expectation, it would be outputted as the final result; In the other hand, the initial tongue body segmentation cannot meet user's expectation, the users have to draw a small part of true contour line, called handraw-line, at failed area.…”