| 摘要 |
PURPOSE: An intrusion detection system using the mining based pattern analysis in a smart home is provided to reduce the overall log data amount and extract a sequential pattern of normal and abnormal behaviors by learning. CONSTITUTION: A user data acquiring part (100) obtains user information for positioning and recognizing a user. An intrusion detection part (200) determines an invasion by performing the mining-based analysis on the user information obtained in the user data acquiring part. The user data acquiring part includes an RFID reader (101), a camera (102), and a communication module (103). The RFID reader reads an RFID tag attached to the user to check user location. The camera obtains a user image for recognizing the user. The communication module delivers the user information obtained by the RFID reader and the camera to the intrusion detection part. The intrusion detection part includes a location estimating part and a position configuring data restoring part. The location estimating part estimates the location of the user based on the user information obtained in the user data acquiring part. The position configuring data restoring part restores a hazardous path by comparing the estimated location information by the location estimating part and a preset virtual position path. [Reference numerals] (100) User data acquiring part; (101) RFID reader; (102) Camera; (103) Communication module; (200) Intrusion detection part; (300) Electronic appliance; (400) Alarm part |
