Abstract: There exist many deficiencies in current video surveillance system such as high expenditure, information omitting, data analysis difficulty, long response. This study proposes a design scheme of intelligent alarm system with the main technology of motion detection, image compression and WEB server pre-fetch which improves the algorithm based on differential image motion detection, controls the costs and reduces false positives. The system features as low cost but fast response with improvement of Loeffler algorithm and the quality of image compression while the decreasing of storage space and transmission bandwidth. It also improves the intelligent Wcol prediction algorithm, predicts the page that user will read soon and eventually improves the efficiency of WEB server. Experiments show that: this system reduces the false alarm rate, avoids filtering, has certain adaptability to light changes, reduces or even eliminates the interference of the CCD jitter, can accurately extract the moving object location. The total time of compressing 100 BMP images and completing the USB storage is lower than 48 sec. The performance that the server with pre-fetch module visits the next page at the second time is better than the performance that the server with no pre-fetch module visits the next page at the second time.