Kovacs L, Zimmermann A, Brockmann G, Gühring M, Baurecht H, Papadopulos N A, Schwenzer-Zimmerer K, Sader R, Biemer E, Zeilhofer H F
Department of Plastic and Reconstructive Surgery, Klinikum rechts der Isar, Technical University Munich, Ismaningerstr. 22, D-81675 Munich, Germany.
J Plast Reconstr Aesthet Surg. 2006;59(11):1193-202. doi: 10.1016/j.bjps.2005.10.025. Epub 2006 Mar 9.
Three-dimensional recording of the surface of the human body or of certain anatomical areas has gained an ever increasing importance in recent years. When recording living surfaces, such as the human face, not only has a varying degree of surface complexity to be accounted for, but also a variety of other factors, such as motion artefacts. It is of importance to establish standards for the recording procedure, which will optimise results and allow for better comparison and validation. In the study presented here, the faces of five male test persons were scanned in different experimental settings using non-contact 3D digitisers, type Minolta Vivid 910). Among others, the influence of the number of scanners used, the angle of recording, the head position of the test person, the impact of the examiner and of examination time on accuracy and precision of the virtual face models generated from the scanner data with specialised software were investigated. Computed data derived from the virtual models were compared to corresponding reference measurements carried out manually between defined landmarks on the test persons' faces. We describe experimental conditions that were of benefit in optimising the quality of scanner recording and the reliability of three-dimensional surface imaging. However, almost 50% of distances between landmarks derived from the virtual models deviated more than 2mm from the reference of manual measurements on the volunteers' faces.
近年来，人体表面或某些解剖区域的三维记录变得越来越重要。在记录诸如人脸等活体表面时，不仅要考虑不同程度的表面复杂性，还要考虑各种其他因素，如运动伪影。为记录程序建立标准非常重要，这将优化结果并便于更好地进行比较和验证。在本文介绍的研究中，使用非接触式三维数字化仪（Minolta Vivid 910型）在不同实验设置下对五名男性测试对象的面部进行了扫描。除其他因素外，还研究了使用的扫描仪数量、记录角度、测试对象的头部位置、检查人员的影响以及检查时间对通过专用软件从扫描仪数据生成的虚拟面部模型的准确性和精确性的影响。将从虚拟模型得出的计算数据与在测试对象面部的定义地标之间手动进行的相应参考测量进行比较。我们描述了有助于优化扫描仪记录质量和三维表面成像可靠性的实验条件。然而，从虚拟模型得出的地标之间的距离中，近50%与志愿者面部手动测量的参考值偏差超过2毫米。
