The Cryptosystem of Kinect Posture Recognition and Speech Recognition

• Main Authors: Cheng-Yi Chu, 朱承毅
• Other Authors: Te-Kwei Wang
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/rjvy9j

碩士 === 明志科技大學 === 電機工程系碩士班 === 102 === Traditionally, most access control systems have had a mechanical structure, such as keys and buttons. With the advancement of technologies, the development of intelligent buildings and the demands of the population for intelligent living, technological access control systems, such as radio frequency identification (RFID) technology and genetic DNA identification technology (e.g., fingerprints), have gradually started emerging in our daily lives. However, these technologies could still be copied, damaged or stolen by thieves. Therefore, the safety of access control systems has always been a major issue. Thus, using Windows 7 as the development platform, the present study utilized speech recognition technology and the KINECT posture recognition technology to construct an access control system, which was then integrated with a contactless password lock. In other words, the proposed KINECT speech password lock system did not require a user to carry anything. Instead, the user opens the lock using a voice command. The user then uses his body as the password-inputting medium to make a specific gesture for the system to recognize. Finally, the system will determine whether the user has inputted the correct password. If the correct password has been inputted, the system will open the door. If not, the system will capture a photograph of the user and record the time when the user input the password as a reference. This study demonstrates that the proposed access control system successfully implemented the concept of contactless password locks; in addition, the proposed system had a good recognition rate.