TY - GEN
T1 - A method for face affine registration via rough search and fine registration
AU - Wu, Liyang
AU - Xiong, Lei
AU - Zhang, Sirui
AU - Du, Shaoyi
N1 - Publisher Copyright:
© 2017 AΒociation for Computing Machinery.
PY - 2017/11/7
Y1 - 2017/11/7
N2 - This paper presents a method for face affine registration which only needs a single template, such as using a person's front face to match his side face. First, in order to be robust to outliers caused by facial affine deformations, an improved color feature is introduced to balance the color similarity and shape mismatch rate between the template face image and target face image, and then the face rough search algorithm is proposed based on the color feature. Next, the affine transformation obtained by face rough search algorithm is used as an initial constraint to establish the face shape fine registration algorithm based on the previous affine constraints. Our algorithm succeΒfully solves the problem that face shape is difficult to register in the presence of rotation, scale transformation and noise interference. Experimental results demonstrate that the robustneΒ and accuracy of our algorithm are greatly improved compared with the conventional face affine registration algorithms.
AB - This paper presents a method for face affine registration which only needs a single template, such as using a person's front face to match his side face. First, in order to be robust to outliers caused by facial affine deformations, an improved color feature is introduced to balance the color similarity and shape mismatch rate between the template face image and target face image, and then the face rough search algorithm is proposed based on the color feature. Next, the affine transformation obtained by face rough search algorithm is used as an initial constraint to establish the face shape fine registration algorithm based on the previous affine constraints. Our algorithm succeΒfully solves the problem that face shape is difficult to register in the presence of rotation, scale transformation and noise interference. Experimental results demonstrate that the robustneΒ and accuracy of our algorithm are greatly improved compared with the conventional face affine registration algorithms.
KW - Affine transformation
KW - Face affine registration
KW - Fine registration
KW - Image registration
KW - Rough search
UR - https://www.scopus.com/pages/publications/85045878868
U2 - 10.1145/3158233.3159358
DO - 10.1145/3158233.3159358
M3 - 会议稿件
AN - SCOPUS:85045878868
T3 - ACM International Conference Proceeding Series
SP - 279
EP - 283
BT - Proceedings of 2017 2nd International Conference on Communication and Information Systems, ICCIS 2017
PB - Association for Computing Machinery
T2 - 2nd International Conference on Communication and Information Systems, ICCIS 2017
Y2 - 7 November 2017 through 9 November 2017
ER -