人臉辨識智慧門禁系統創新探討

Abstract

因為智慧安全防護逐漸受到市場的關注，加上物聯網技術日漸普及，配合生物特徵辨識中的人臉辨識發展速度下，本研究結合技術創新理論探討智慧門禁系統結合人臉辨識技術，歸納出五大構面與十八項準則，五大構面分別為人臉辨識技術、系統集成性、系統安全性、系統可靠性和系統易用性。 本研究問卷發放對象為產業應用、學術研究和政府單位領域中，具備人臉辨識、智慧安防或系統開發專業相關背景之 22 位專家。利用多層次決策分析法探討人臉辨識智慧門禁系統創新架構中，構面之間與準則之間的重要排序性關係，藉由結論評估未來發展策略以利於智慧門禁的產業發展。 本研究結果得知五大構面中以系統安全性為優先考量的重點屬性。五大構面底下準則之優先考量屬性分述如下: 一、人臉辨識技術構面下為正確辨識率。二、系統集成性構面下為數據集成。三、系統安全性構面下為隱私保護。四、系統可靠性構面下為平均無故障時間。五、系統易用性構面下為有效性。本研究結果可提供開發廠商或設計者研擬未來適合人臉辨識智慧門禁系統創新策略方向。

Because smart security protection has gradually received market attention, coupled with the increasing popularity of Internet of Things technology, and the development speed of facial recognition in biometric recognition, this research combines the theory of technological innovation to explore smart access control systems combined with facial recognition technology, and summarizes five major Aspects and eighteen criteria, the five aspects are facial recognition technology, system integration, system security, system reliability and system ease of use. This research questionnaire is distributed to 22 experts with backgrounds in facial recognition, smart security, or system development in the fields of industrial applications, academic research, and government units. Use the multi-level decision analysis method to explore the important ordering relationship between the aspects and the criteria in the innovative architecture of the facial recognition smart access control system, and use the conclusion to evaluate future development strategies to facilitate the development of the smart access control industry. The results of this study have learned that the key attributes of the five dimensions that prioritize system security are considered. The priority attributes of the criteria under the five dimensions are as follows: 1. The correct recognition rate under the facial recognition technology dimension. 2. Data integration is under the system integration dimension. 3. Privacy protection under the system security aspect. 4. The system reliability is the mean time between failures. 5. The effectiveness of the system is under the aspect of ease of use. The results of this research can provide developers or designers with the direction of future innovation strategies for smart access control systems suitable for facial recognition.