Summary: | 碩士 === 國立臺北教育大學 === 資訊科學系碩士班 === 103 === And now most people have mobile devices, almost all of the mobile devices will be equipped with the lens, such as smart phones and tablets. Camera applications will be more changes in the future. Therefore, facial recognition issues more and more attention, and facial recognition easier in the case of changing light under affect face recognition rate. This thesis proposed a method for improving the light change problem. First, the database of facial recognition to do face pre-treatment improves before the light changes, then use the PCA algorithm, prior training data of database, get facial image eigenvalues, and compared with the input image to find out closer the face. The experimental results show that proposed the algorithm in this paper can improve the lighting changes, and enhance the face recognition rate.
|