Accuracy of Smartphone-Based PhotogrammetricThree-Dimensional Facial Models for Anthropometric Measurements

Objective:To evaluate the accuracy of three-dimensional (3D) facial models reconstructed from smartphone photographs using photogrammetry software, compared with direct anthropometric measurements as the gold standard. Methods:Fifty participants (25 males, 25 females; aged 18 years or older) were recruited. Fifteen linear facial measurements were obtained directly using a digital caliper and compared with measurements derived from 3D facial models reconstructed from 25 overlapping smartphone images using 3DF Zephyr Free (v7.531). Angular accuracy was assessed using standardized LEGO bricks (31.8 mm × 15.8 mm×9.6mm)attachedtotheforehead and cheek,with 8 angles measuredper brick and compared against the reference angle (90°). Paired sample t-tests were used for statistical comparisons, with clinical acceptability thresholds set at ≤2 mm for linear and ≤2° for angular measurements. Results: For linear measurements, 14 of 15 measurements showed no statistically significant differences from direct anthropometry (p > 0.05); only Sn-Pog showed a significant difference (p = 0.014), with a mean deviation of 0.51mm. Theoverallmean absoluteerror was0.44 mm.All 16angular measurements fell within the ≤2° threshold (range: 0.10°–1.66°), with cheek errors (0.13°–1.66°) larger than forehead errors (0.10°–0.31°). Conclusions: Smartphone-based photogrammetry using 3DF Zephyr Free provides clinically acceptable accuracy for 3D facial anthropometric measurements. This technology offers a costeffective, accessible, and radiation-free alternative for orthodontic facial analysis.