A method for measuring hairline length and discriminating hairline recession grades based on the BiSeNet model

Yuhua Ai,Guoliang Wei,Junke Wu,Shuting Liu,Bo Chen,Zhenyu Zhang,Qimin Li
DOI: https://doi.org/10.1088/1361-6501/ad886e
IF: 2.398
2024-10-20
Measurement Science and Technology
Abstract:Hair plays an important role in a person's appearance. According to a survey by the World Health Organization, approximately 70% of adults have scalp and hair prob lems. Doctors currently make hairline recession diagnoses based on hair loss criteria, but this approach is subjective. This paper proposes a novel method for objectively assessing hairline recession grades. First, the Bilateral Segmentation Network model (BiSeNet) is utilized to obtain a facial segmentation image. Second, this paper utilizes the connected components method to improve the facial segmentation results. Next, the labeling key points method is used to extract part of the features of the eyebrow and facial region and calculate the related values. Finally, the judgment of hairline length and hairline recession grade is realized by combining these features with camera calibration. In this paper, front-face images of 50 volunteers were collected for hair line recession grade determination. The judgment results of expert doctors on hairline length and hairline recession grade were compared with the judgment results of this method. The results showed a 1.3 cm difference in the average length of the hairline and about 80% similarity in hairline recession grade judgments. In conclusion, using machine vision methods to measure the height of the hairline provides objective and repeatable results.
engineering, multidisciplinary,instruments & instrumentation
What problem does this paper attempt to address?