Symmetric Plane Detection and Symmetry Analysis from A 3D Point Cloud Data of Face

• Published in: International Conference on Control, Automation and Systems (Online) pp. 402 - 406
• Main Authors: Hosoki, Daisuke, Kamiya, Tohru, Kimura-Nomoto, Namiko, Okawachi, Takako, Nozoe, Etsuro, Nakamura, Norifumi
• Format: Conference Proceeding
• Language: English; Japanese
• Published: Institute of Control, Robotics, and Systems - ICROS 13.10.2020
• Subjects: Cleft Lip; Faces; Feature extraction; Iterative Closest Point; Lips; Nose; Point Cloud; Shape; Surgery; Symmetric Analysis; Three-dimensional displays
• ISSN: 2642-3901
• DOI: 10.23919/ICCAS50221.2020.9268262

Cleft lip occurs in about 1 in 500 people and is a congenital anomaly that occurs when the lips are not completely formed in the process of fetal face fusion. The treatment of the cleft lip is performed for the purpose of forming bilaterally symmetrical lips and external nose. However, it is based on a criterion that depends on subjectivity, quantitatively determining the degree of symmetry of the surgical site is required. In this paper, we propose a method for detecting the face that serves as the symmetry basis of the face to analyze the symmetry of the surgical site, and an index that indicates the symmetry. In the proposed method, after detecting the facial organs as points in the 3D point cloud data of the subject's face, we align the mirror image inversion excluding the region from upper lip to the tip of the nose where the effect of shape change due to cleft lip is remarkable with the original point cloud. Next, the symmetric reference plane is set by finding a plane that bisects vertically between one point in the original point cloud and the corresponding point in the mirror image inversion. As a result, the target reference plane is detected with high accuracy. Furthermore, as a result of applying the symmetry analysis method based on the target reference plane, it was possible to visually represent the left-right asymmetric region.