基于ARM的嵌入式图像监控系统硕士学位论文.docxVIP

II Study of an Embedded Image Monitoring System Based on ARM Discipline: Control theory and control engineering Student signature: Supervisor signature: ABSTRACT With the rapid evolution of digital image, embedded computer and communication technique, embedded image monitoring system becomes a new researching interest. However, a large number of pointless pictures are transmitted back to the monitoring center, which increases communication quantity and surveillance load; generally, the image’s definition in remote surveillance is low because of its limited transmission channel. An improving way is moving object detection in the front of system, and snatching a high definition picture when motion happens and then transferring back to the monitoring center. This method can settle the contradiction between definition and transferring speed. In the dissertation, a kind of embedded image monitoring system based on ARM is discussed. The system has the functions of motion detection and picture transmission based on GPRS, and improves the flexibility of setting and the definition of abnormal picture. The system takes LPC2210 based on ARM7 core as its central processor, and embedded OS μC/OS-II as its software platform. Moving object detection adopts a method of dynamic difference background image, whose principle is to calculate difference picture among the current frames and background images and take the motion adjudging threshold which owns maximum information entropy to determine moving area; and dynamically updates background by using Kalman filter. Thus, the system realizes the judgment of the moving area under the complicated changing background. The monitoring system adopts embedded Web server technology to send warning information and on-the-spot picture to the remote monitoring center. The wired or wireless remote transmission channel (Ethernet