To improve the reliability of security monitoring results and achieve intelligent monitoring of buildings, research and design of building security intelligent monitoring methods were carried out based on sensor data fusion and scale invariant feature transform(SIFT) feature matching. Cameras, infrared detectors, vibration sensors were installed, building security data was collected, weighted average method was introduced to fuse and compress preprocessed data with point cloud data. Color images was converted into grayscale images, key points (feature points) and corresponding descriptors were extracted, and automatic stitching of building security monitoring images was designed based on SIFT feature matching. In the spliced video frames, the inter frame difference method was used to identify moving targets, and a threshold was set to distinguish between foreground(moving targets) and background, achieving tracking and intelligent monitoring of video moving targets. The comparative experimental results show that the designed method has a good effect in practical applications. This method can accurately identify all characters appearing in the monitoring interface, meeting the requirements of intelligent monitoring for building security.

sensor data fusion  /  point cloud data compression  /  intelligent monitoring  /  security and protection  /  buildings  /  scale invariant feature transform (SIFT) feature matching