Robust ECG Biometric Identification By Using Post Exercise ECG

• Main Authors: Shan-chun Chang, 張善鈞
• Other Authors: Tsu-wang Shen
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/03903648363440890011

碩士 === 慈濟大學 === 醫學資訊學系碩士班 === 100 === There are always needs for human identity authentication. No matter what kind of personal services or web services nowadays, privacy and security of personal data are essential. Because medical informatics is a fast growing area, the security of personal identity becomes very important. An efficient and safe authentication mechanism is highly focused technology for personal ID protection applications. ECG biometrics has several unique advantages, including difficulty for stealing/counterfeiting and authentication with liveness check. However, heart rate (HR) changes are the one of the most intractable problems in real-world ECG biometric applications. Currently, specific rules or patterns of ECG morphology changes according to heart rate changes are still unknown. The performance evaluations of heart rate (HR) changes of substantial ECG biometric system are lack on pervious literatures, which makes this research became more challenge. 50 people were involved in the study and total of 300 exercise ECG patterns from those subjects were collected. After combined with several methods, the recognition rates at the first 4 heart rate (HR) stages is up to 92% and overall system identification rate is up to 82%. At the end of article, the limitations will be discussed, and a possible solution is provided.