Academic Research of Biometrics in Korea

Transcription

1 Academic Research of Biometrics in Korea Ho Gi Jung Academic research of biometrics in Korea can be summarized into two groups: a research group belonging to biometric engineering research center (BERC) and laboratories researching biometrics as one of application fields of pattern recognition or computer vision. 1. BERC BERC (homepage was founded by a Korean governmental organization, the Korea science and engineering foundation (KOSEF), in 2002 at Yonsei University, Seoul in order to develop core new technologies for biometrics, technologies for biometrics information security, and multi-modal biometrics at a distance. Its director is Prof. Jaihie Kim and its members are from several universities, 17 professors and more than 100 experts. It made memorandum of understandings (MOU) with 5 domestic institutes such as National Institute of Scientific Investigation (NISI), Korea Internet & Security Agency (KISA), and Electronics and Telecommunication Research Institute (ETRI). For international corporations, it made MOUs with 7 foreign institutes such as Biometrics Security Research Group. Additionally, it has close relationships with 13 domestic companies specializing in biometrics. Studies on biometrics core new technologies are under the control of Prof. Jaihie Kim of Yonsei University (homepage The goal has been made concrete by four subjects. 1) Contactless fingerprint recognition, under the control of Prof. Jaihie Kim, aims at development of contactless type equipment for acquisition of 5 fingerprints, development of contactless type image enhancement and image fusion. It also includes studies on compatibility of contactless type fingerprint image with contact type one. 2) Non-intrusive iris recognition, under the control of Prof. Moongi Kang of Yonsei University (homepage aims at restoration of super-resolution iris/face images using low

2 resolution continuous images, development of low cost long distance iris recognition system for non-intrusive iris recognition, and mobile iris recognition system robust to sunlight. 3) 3D face recognition, under the control of Prof. Kwanghoon Sohn of Yonsei University (homepage aims at 3D face modeling using 2D face images, development of 3D face data acquisition equipment, development of 2D & 3D face recognition technologies, development of long distance face recognition technologies in fish-eye monitoring camera environment, and development of active appearance model (AAM) based high-speed face characteristics extracting technologies which are strong against changeable internal & external environment. In particular, Prof. Daijin Kim of Pohang University of Science and Technology (POSTECH) (homepage made a significant achievement in the field of 2D & 3D face recognition. 4) Studies on speaker recognition, under the control of Prof. Mooyoung Kim of Sejong University (homepage aims at development of speaker recognition technologies in different environments. Studies on biometric information security technologies, under the control of Prof. Kang Ryoung Park of Dongguk University (homepage has been conducted to build a high level biometrics system which provides better security and recognition performance by preventing leak of personal biometric information which may occur in the biometrics system. It includes biometric encryption, changeable biometrics, and identification of fake or altered biometric information. The goal was followed by two subjects. 1) Studies on fake fingerprint identification, under the control of Prof. Hakil Kim of Inha University (homepage include understanding of static images quality characteristics (static feature) about real personal fingerprint images and fake ones, dynamic feature based fake fingerprint recognition according to user interaction, and performance improvement for practical applications. 2) Biometric encryption & changeable biometrics, under the control of Prof. Kang Ryoung Park, include security protection of biometric information by encryption biometric information & biometric unique key creation from iris, face, and fingerprint.

3 Studies on multimodal biometrics technologies at a distance, under the control of Prof. Kar Ann Toh of Yonsei University (homepage include studies on long distance biometrics based on human gait to overcome limits of personal biometrics and studies on multimodal biometrics technologies of profile or front facial biometric information at a distance. 2. Independent Research Laboratories Many academic laboratories specializing computer vision and pattern recognition treat face recognition, and gait recognition as one of applications. Prof. Joonki Paik of Chung-Ang University (homepage has been researching enhancement of compressed image & video, contrast enhancement & color image processing, model-based object tracking, multi-modal object tracking, multiple camera tracking, geometry-based multiple camera surveillance, image communication, range data smoothing, 3D scene & object modeling, 3D data segmentation using super-quadric, digital intermediate, and content wrapper technology. In the field of biometrics, he developed 2D & 3D face recognition, and gait recognition. Prof. Seong-Whan Lee of Korea University (homepage has been developing various vision systems including gesture recognition, sign language recognition, intelligent video surveillance, and human behavior analysis. He also has been developed brain image analysis, brain-machine interface, cognitive neuroscience, and affective computing. In the field of biometrics, he combined real time face detection & tracking with 3D face reconstruction and recognition. Prof. Sang Uk Lee of Seoul University (homepage has been researching various basic computer vision problems including energy minimization, color & illumination, colorization, segmentation, and medical imaging. In the field of biometrics, he developed face recognition using face-arg (attributed relational graph) matching.

4 Prof. Myung-Geun Chun of Chungbuk University (homepage developed contourlet and PCA (principal component analysis)- based face recognition, Gabor wavelet and fuzzy LDA (linear discriminant analysis)-based iris recognition, and biometric features and responsible person information hiding by watermarking technique. One special approach to face recognition has been conducted by brain science research center (BSRC), which was established at Korea Advanced Institute of Science and Technology (KAIST) with a goal to build an intelligent machine with a brain modeled after human s through understanding, imitation, and application of neural information processing mechanism which has been studied in two separate fields, brain science and biomedical brain. In the sub-project, object vision research and application development based on brain information processing, neural network-based face recognition was developed by Prof. Seungyang Bang of POSTECH (homepage Besides BSRC, similar researches are conducted by a few laboratories. For example, Prof. Dae-Seong Kang of Donga University (homepage developed a learning algorithm for face recognition using brain modeling of Hippocampus. Contrary to face recognition, iris recognition and fingerprint recognition is regarded as peculiar to biometrics and researched by specialized research groups. Prof. Sung-Bae Cho of Yonsei University (homepage developed soft computing-based applications including artificial agent, smart phone, ubiquitous, and robots. In the field of biometrics, he developed image enhancement for fingerprint and localized model support vector machine (SVM)-based fingerprint recognition. Prof. Yoosek Bae of Korea Polytechnic University (homepage researched fake fingerprint detection. There are a few independent laboratories researching biometric system-related problems. Prof. Pil Joong Lee of POSTECH (homepage has been

5 researching cryptography and applied it to biometric template protection. Prof. Yongwha Chung of Korea University (homepage has been researching fingerprint recognition, speaker recognition, and face recognition. In particular, he researched cancelable biometrics. Prof. Young-Bin Kwon of Chung-Ang University (homepage has been researching standardization of biometrics.