Related Experiment Video
Updated: Aug 7, 2025
10:23
Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans
Published on: September 8, 2023
2.9K
The Anthropometric Measurement of Nasal Landmark Locations by Digital 2D Photogrammetry Using the Convolutional
Nguyen Minh Trieu1, Nguyen Truong Thinh1
1College of Technology and Design, University of Economics Ho Chi Minh City-UEH, Ho Chi Minh City 72516, Vietnam.
Diagnostics (Basel, Switzerland)
|March 11, 2023
Summary
This study introduces a Convolutional Neural Network (CNN) model for automatic facial landmark detection. The system achieves high accuracy in anthropometric measurements, offering a low-cost solution for facial analysis.
Area of Science:
- Medical Imaging
- Computer Vision
- Anthropometry
Background:
- Manual facial landmark annotation is labor-intensive and requires expert knowledge.
- Convolutional Neural Networks (CNNs) have shown significant advancements in image analysis tasks like segmentation and classification.
- Accurate facial anthropometry is crucial for aesthetic procedures like rhinoplasty.
Purpose of the Study:
- To develop and evaluate a CNN-based model for automated extraction of facial landmarks.
- To implement an automatic system for anthropometric measurements using facial landmarks.
- To assess the accuracy and stability of the proposed automated measurement system.
Main Methods:
- A CNN model was trained to learn and recognize facial landmarks based on medical theories.
- The system processed images from frontal, lateral, and mental views for comprehensive analysis.
- Automated measurements included 12 linear distances and 10 angles.
Main Results:
- The CNN model demonstrated effective landmark detection capabilities.
- The automated system achieved a normalized mean error (NME) of 1.05.
- Average errors were 0.508 mm for linear measurements and 0.498° for angle measurements.
Conclusions:
- The study successfully proposed a low-cost automatic anthropometric measurement system.
- The system exhibits high accuracy and stability in facial landmark detection and measurement.
- This automated approach can significantly reduce the time and cost associated with facial analysis.
