Three-dimensional Facial Images Research Articles

Synthesis and transformation of three-dimensional facial images.

In this article, we enhance and extend 2-D methods by using texture-mapped laser-video depth data. Methods for the synthesis of new three-dimensional (3-D) facial images from a database of texture-mapped laser-scanned surfaces are described. Four methods are used to create new facial images. The first constructs average facial images from a sample of faces. The second exaggerates the difference between a subject and an average to create a caricature. The third method captures the difference between two averages, which can then be applied to individuals for statistical transformation. Finally, principal component analysis is used to find the major axes of facial variation in a sample, allowing the construction of completely novel faces from a small vector of values.

Three-Dimensional Facial Anthropometry Using a Laser Surface Scanner: Validation of the Technique

Three-dimensional measurement and characterization of facial surface anatomy are fundamental to the objective analysis of facial deformity. However, existing clinical tools are inadequate. Recent innovations in laser scanning technology provide a potentially useful technique for accurate three-dimensional documentation of the face. The purpose of this study is to evaluate the reliability of interactive anthropometric landmark localization based on digitized three-dimensional facial images and to identify sources of error associated with the technique. Three-dimensional surface data were acquired using a commercially available laser scanning device (Cyberware 3030RGB digitizer), and all subsequent anthropometric analyses were performed interactively on the computer monitor. Four experimental conditions were studied, with 10 observations for each condition. A stable anthropomorphic model with prelabeled anatomic landmarks was scanned repeatedly under varying conditions of head inclination and position within the scanning gantry to determine the effect of these variables on the reliability of the technique. The scanning protocol was then repeated with the labels removed to evaluate the reliability of interactive localization of anthropometric landmarks on a digitized three-dimensional image. Optimal results were obtained with the head positioned in the center of the scanning gantry and with the Frankfort plane elevated 10 degrees from the horizontal. Under these circumstances, all 22 labeled landmarks were visualized and the variance in landmark localization was less than 0.6 mm in the x (horizontal), y (vertical), and z (depth) dimensions. Varying head position or inclination caused significant degradation of the digitized image. The variance of interactive landmark localization was analyzed in three dimensions. The reliability and the spatial orientation of variability were determined for each anthropometric point. These findings have direct implications for the clinical adaptation of this diagnostic tool for quantitative evaluation of facial surface anatomy.