90-Degree Angle Side-View Facial Matching

• Main Authors: Yen-Yu Chou, 周彥宇
• Other Authors: Sendren Sheng-Dong Xu
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/74wwcq

碩士 === 國立臺灣科技大學 === 自動化及控制研究所 === 107 === In this thesis, three maching methods in image processing, including histogram comparison, template matching, and average hash methods, are used to explore the facial maching of 90-degree side-view face images. The FEI face database is used for this research. First, we pick a 90-degree side face image that includes a total of 160 images for men and women, ten of which are black and white. Then, we scale the image to 256 x 256 pixels and compare the remaining 159 images to the first photo. Finally, experimental results show that by using the average hashing method, only 4 pictures are misjudged, and the performance is less affected by the colors of the test images. Its performance is the best among the three methods.