Suppr超能文献

计算机辅助面部重建系统。

The computer-aided facial reconstruction system.

作者信息

Miyasaka S, Yoshino M, Imaizumi K, Seta S

机构信息

First Medico-Legal Section, National Research Institute of Police Science, Tokyo, Japan.

出版信息

Forensic Sci Int. 1995 Jun 30;74(1-2):155-65. doi: 10.1016/0379-0738(95)01744-4.

Abstract

A computer imaging system was introduced into the facial reconstruction process. The system, which consists of the image processing unit for skull morphometry and the image editing unit for compositing facial components on the skull images, was an original construction. The image processor generates the framework for building a face onto the digitized skull image. For reconstructing a facial image on the framework, several possible data sets of facial components suitable for the skull morphology are selected from the database by operating our original application software. The most suitable cutout samples of facial components are pasted up over the framework in accordance with the anatomical criteria. The database of facial components consists of 24 contours, 18 eyes, 9 eyebrows, 27 noses, 9 lips and 16 hairstyles. After provisional reconstruction, the facial image is retouched by correcting skin colors and shades with an 'electronic painting device'. The resulting image is a great improvement on images made by the conventional clay and drawing method, both in the operational aspect and in the flexibility of creating multiple versions. The present system facilitates a rather objective and rapid approach and allows us easily to generate a range of possible faces. The computer-aided facial reconstruction will lead to an increase in chances of positive identification in practical cases.

摘要

一种计算机成像系统被引入面部重建过程。该系统由用于颅骨形态测量的图像处理单元和用于在颅骨图像上合成面部组件的图像编辑单元组成,是一种原创构造。图像处理器生成在数字化颅骨图像上构建面部的框架。为了在该框架上重建面部图像,通过操作我们的原创应用软件,从数据库中选择几个适合颅骨形态的面部组件可能数据集。根据解剖学标准,将最合适的面部组件剪裁样本粘贴在框架上。面部组件数据库包括24种轮廓、18只眼睛、9种眉毛、27个鼻子、9种嘴唇和16种发型。初步重建后,使用“电子绘画设备”通过校正肤色和色调对面部图像进行修饰。所得图像在操作方面以及创建多个版本的灵活性方面都比传统的黏土和绘图方法制作的图像有了很大改进。本系统有助于采用一种相当客观和快速的方法,并使我们能够轻松生成一系列可能的面部图像。计算机辅助面部重建将增加在实际案例中进行正面识别的机会。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验