CAJ | 학술논문

采用层级处理模式，提出了一种新的人脸光照归一化方法，先对图像光照进行自动判定，决定采用何种方法进行预处理，减小不同光照下人脸灰度的差异。再根据Gamma校正算法的特点，设计了一种更适应人脸图像校正的参数计算方法。本方法在样本归一化后，使得图像的本真信息不丢失，而且提高了图像对比度。大量的实验表明，该方法可以有效提高人脸识别功能的适用范围，能够在更远距离或复杂光照下进行识别。
채용층급처리모식，제출료일충신적인검광조귀일화방법，선대도상광조진행자동판정，결정채용하충방법진행예처리，감소불동광조하인검회도적차이。재근거Gamma교정산법적특점，설계료일충경괄응인검도상교정적삼수계산방법。본방법재양본귀일화후，사득도상적본진신식불주실，이차제고료도상대비도。대량적실험표명，해방법가이유효제고인검식별공능적괄용범위，능구재경원거리혹복잡광조하진행식별。
A new face illumination normalization method has been proposed based on hierarchical processing model to judges automatically the illumination of image ,and to decides w hich method is adopted for pre‐treatment ,so as to reduce the face gray differences under different illumination .According to the charac‐teristics of Gamma correction algorithm ,a more suitable method has been designed to calculate the correc‐tion parameters .After normalization by this method ,the facial features will not been lost ,and will im‐prove the contrast of image .A large number of experimental results show that this method can effectively improve the adaptive scope of facial recognition ,and can work under the conditions of longer distances and more complex illumination .