The system can image the way outsiders invasions, fires and other indoor anomalies automatically
determine. Its intelligence is mainly reflected: the system software on the state of the environment, self-study,
so that it can adapt to background changes in the environment, critical state, which can automatically start the
state tracking and monitoring to accurately determine the results of multiple images. Background subtraction
and interframe difference method, based on the movement threshold binding assay. The target detection
process including: image preprocessing, quantization function, motion detection (or a small target motion
tracking and detection), the abnormal situation of discrimination and processing.
When the system receives a new image to the camera sent the first pre-processing and noise removal, to get
a binary image that can quantify the movement of objects or abnormal areas, and then proceed to the analysis
and judgment of the abnormal state. The system uses two kinds of abnormal quantify function: Based on the
global bitmap handle and random testing, processing, and then, according to the exception to quantify the
function pointer to a specific follow-up operation, For to enter the alarm exception handler, then send an
alarm signal to send MMS messages, and store the current video, For to enter the critical exception tracking,
start the critical exception tracking program, otherwise the background to replace the judgment and wait to get
under a monitoring image. Transmission distance without obstacles within 80 m, the transmission of data
packet loss rate of less than 1% are adequate to meet the communication needs of the normal home
environment, and the received power is very low, very suitable for family use ZigBee modules for
communication between rate of 0% packet loss when the transmission distance is 30 m, through the loss of
two water shutoff walls only 1% through a water shutoff walls, fully able to meet the requirements of the
system design.