摘要： 利用图像的边缘检测与提取技术自动识别照片的人像位置，判别并修正背景颜色、皮肤色彩范围、灰度偏差等人像信息，使之符合居民身份证制证用数字相片的样式规定和要求。这一技术在应用中取得了令人满意的成绩，尤其是其对特殊人群的照片处理有着非常显著的效果，是数字照片合格率准确性、稳定性和有效性的有利保证，也使得人像信息检测的自动化、智能化、批量化成为可能。

关键词: 边缘检测, 轮廓提取, 边界修补, 灰度阶跃, 差影法

Abstract: The innovated edge detecting and extracting technology helps accurately position objects in the photograph, thus able to automatically distinguish and alter the background color of the photograph, correct the object’s skin color, adjust the tolerance level, etc. Ensuring the photograph fulfills the resident identity card digital photograph specifications and requirements. When the technique is applied, not only are the results satisfactory, but also significant differences can be observed especially for some special groups of people after their photographs have been processed. This innovation allows the digital portrait photograph creation to be ever accurate, stable and efficient. In addition, because of its ability to be fully automated and highly intellectualized, photographs are able to be processed in large batches.

Key words: Edge detection, Contours extracting, Edge retouch, Gray step, Difference image methods