Keypoint Descriptors in SIFT and SURF for Face Feature Extractions

• Main Authors: Suk, Ting Pui, Minoi, Jacey Lynn
• Format: Proceeding
• Language: English
• Published: Springer Verlag 2018
• Subjects: T Technology (General)
• Online Access: http://ir.unimas.my/id/eprint/20292/1/Keypoint%20Descriptors.pdf; http://ir.unimas.my/id/eprint/20292/; https://link.springer.com/chapter/10.1007/978-981-10-8276-4_7
• Institution: Universiti Malaysia Sarawak

The last decade, numerous researches are still working on developing a robust and faster keypoints image descriptors algorithm. In this paper, we will review a few complex keypoint descriptor approaches that are well-known and commonly used in vision applications, and they are Scale Invariant Feature Transform (SIFT) and Speed-up Robust Features (SURF). These methods aim to make the descriptors faster to compute and robust to scale, rotation and noise. We will the results of the experiments on face image data. The extracted keypoints and the regions of interest are analysed and compared against the corresponding facial features. The results have shown SIFT outperformed SURF in terms of speed while the extracted keypoints using SURF descriptors are mainly located on the corners and distinct facial features. © 2018, Springer Nature Singapore Pte Ltd.