An Algorithm for Extraction of Near Infrared Sublingual Veins

Abstract

Sublingual vein is one important diagnostic measurement in Tongue Diagnosis. Segmentation method of sublingual veins from the near infrared sublingual images is proposed in this paper. According to the characteristic of the captured near infrared sublingual images, the radial projection method based on watershed algorithm is firstly applied to acquire the entire contour of the sublingual region, and then Dynamic thresholding is utilized to complete the binarization of the obtained effective sublingual region, and further confirm the candidate sublingual veins regions by considering the intensity feature of near infrared sublingual vein image. Finally, region growing is iteratively performed, which is restricted by original intensity feature of the near infrared sublingual image, to trace the contour of sublingual veins. Experimental results reveal that, the proposed methods did effectively isolate the sublingual veins and the surrounding tongue proper, and precisely trace the boundary of the sublingual veins.