CAJ | 학술논문

通过人脸图像对年龄进行识别具有很大的挑战。年龄识别会受到诸多干扰因素的影响，如性别、健康、生活方式等等。本文设计了一种基于主成份分析人脸图片的年龄识别方法，选用 FG-NET 数据库的图片作为测试样本，挑选了测试集和训练集。论文对选择的图片作直方图均衡化，直接计算图片的欧氏距离并判别近邻，之后，对图片进行人脸提取矫正等预处理后进行测试，最后使用本文方法对人脸提取后的图片进行测试和实验。
통과인검도상대년령진행식별구유흔대적도전。년령식별회수도제다간우인소적영향，여성별、건강、생활방식등등。본문설계료일충기우주성빈분석인검도편적년령식별방법，선용 FG-NET 수거고적도편작위측시양본，도선료측시집화훈련집。논문대선택적도편작직방도균형화，직접계산도편적구씨거리병판별근린，지후，대도편진행인검제취교정등예처리후진행측시，최후사용본문방법대인검제취후적도편진행측시화실험。
The image-based face-age evaluation is a significant technology, which involves many real-world appli-cations, and there remains a great challenge through age evaluation. Age evaluation is influenced by various factors, such as gender, health, and lifestyle. This paper designs a PCA-based face-age evaluation method and selects images from FG-NET database as the test sample, part of which is chosen as test list and train list. Then the pretreatment (his-togram equalization) is done to all chosen images, and the Euclidean distance is calculated to discriminate the nearest neighbor. Then the author extracts the face from all images above and test again. After that, the method proposed in this paper is used to test the extracted images.