A Research Of Digital Image Management System

Real-time video surveillance system is an important branch of image processing. Nowadays,it plays an important role in real life and production. As people’s requirements on the quality ofmonitoring image are increasing, traditional analog cameras are unable to meet the needs ofpractical applications any more. With the rapid development of digital multimedia and network,video surveillance systems with digital camera gradually become the mainstream products ofmarket. The image management system proposed in this paper is also based on digital cameras.This system is responsible for transferring and compressing the images captured by digital camera.According to the requirement, the whole system is divided into two parts: digital cameras andthe server. The core processor of digital camera is ARM, which is responsible for shooting the realscene and sending sample images to the server through network instantly. On the other side, theserver terminal gets the complete image sequence by analyzing the received network packets anddisplays the image sequence, and then, server compresses and stores the image data in differentways according to needs.During developing the system, the main problem is the large amount of data, high complexityof the procedure and the real-time requirement of system. As is mainly reflected in this system, thesystem is required real-time feedback of the images captured by the camera and needs to compressand store them in a short time. In addition, because the images are all24-bit true color images andcamera has a high sampling frame rate, system has to process large amount of data. This is a bigchallenge for processing speed.This paper focuses on transmission and management of digital images. The large amount ofdata and the processing speed are mainly reflected in the following aspects: Firstly, the resourcesof camera’s processor and storage space are limited as while as a lot of data reading and writingoperations are needed in capturing and sending images, this may influence the camera’s performance. Secondly, when sending the image date to server through network, the networktransmission speed is difficult to control because of the complex network environment.Meanwhile, data packets may lose or arrive in incorrect sequence. Thirdly, a high complexitycalculation is needed in the compression process of image data. It is difficult to satisfy therequirements for both the high compression ratio and image quality. Therefore, in theimplementation, the system should reduce the loads of camera’s processor, improve the reliabilityand speed of network transmission and realize the fast image compression. At the same time, thepractical application scenarios must be taken into account to achieve real-time image processing.For the problems of digital image management and transmission, a useful solution with goodperformance is proposed in this thesis. Considering the limitations of camera’s device, only thesampling and sending digital images are executed by digital cameras. By using the DMAcontroller, it enables to avoid the frequent data reading and writing operations from disturbing thenormal work of ARM processor. When sending digital images, TCP is used as the transport layerprotocol to ensure the reliability of data transmission. Meanwhile, a lightweight network protocolstack is used to improve transmission efficiency. According to the requirements, image data can becompressed and stored in different ways. To achieve the high quality and efficiency imagessequence compression, the H.264coding standard is used. Finally, this system achieves theprocesses of digital images’ acquisition, transmission, analysis, display, compression and storage,which lay the foundation for the analysis and recognition of the subsequent monitoring images.