中文说明：
ORL人脸图像库，共40人，每人10幅图像，其中每人的前5幅作为训练样本，后5幅作为测试分类样本，统计正确分类率。分类准则为最近邻规则。 真实的图像尺寸为112x92，列向量堆积对应人脸库矩阵的每一列。
English Description：
ORL face image database, a total of 40 people, 10 images per person, of which the first 5 images per person are used as training samples and the last 5 images are used as test classification samples to count the correct classification rate. The classification criterion is the nearest neighbor rule.