Aiming at solving the problems of low detection efficiency and high false alarm rate of moving targets detection algorithm in embedded alarm system because of the interference of noise

a double-threshold background difference algorithm based on object characteristic is proposed.Based on dynamic threshold background difference algorithm and in accordance with object characteristics

two thresholds are used to detect targets by regulating the proportion of moving targets over the whole image

and update background model in real time.This method can correctly segment the image and effectively remove the noise.The experimental results show that this algorithm with low computation quantity and good robustness is suitable for the strong real-time embedded alarm system.