The Research On Network Monitor System Based On Picture

In spite of the growing popularity of embedded video surveillance technology, this paper describes an image monitoring system. Although image expresses less information compared with video, its transmission saves data bytes and costs less money, and leads to good network environments. More importantly, this type of device has simple architecture and can be put into the sleep mode most of time, thus the power is saved and possibly the battery(even no special battery) can be used for a prolonged time of monitoring. Image is sufficient to express what happens in some scenarios. For example, it is valuable to send pictures of an uninhabited room or visitors with a battery-powered device. Leaving home while keeping the mains power(220V) off,it is a common habit for many Chinese people.This paper mainly designs an embedded image monitoring system based on ENC28J60.The heart of the hardware is a medium performance microcontroller STM32F103RCT6 produced by ST corporation,with a camera produced by OmniVison corporation to collect photos. The ENC28J60 is used to interface with the Ethernet. The drivers of the parts and the sleeping mechanism are designed. The principle of μIP protocol and the transplanting method of Contiki operating system are also depicted. Lastly,the embedded system can transmit real-time data through the OneNet platform or OneNet device cloud over the Internet. The administrator can login into the system and do the monitoring remotely.This paper also researches a scheme of wireless image monitoring system based on the GS1011 M WIFI module of low-power consumption. With the above architecture, the system accesses the network by GS1011 M. It can simply be controlled by “AT” commands and has excellent performance in low power consumption.Clearly programming on bare ARM machine and wireless communication is the most significant advantage of this system.After that,this paper also implements a clipping algorithm of JPEG pictures to cut down the data size. To obtain a smaller size with less time and space, this paper adopts the Huffman table recommended by JPEG and the data outside the clipping window is deleted. As a result, the size of an image is decreased significantly which improves the performance of our system in reliability and real-time.