The Application of Embedded Teaching and ...

4 downloads 327 Views 987KB Size Report
Relational Markov networks (RMNs), which extended from Markov random fields (or ... The hardware structure of the embedded monitoring system is shown Fig.1. ... with development board, Camera driving uses Linux built-in ov511 chip .... used to send commands and start serial GPRS module to mobiles alarm messages.
The Application of Embedded Teaching and Experiment Platform: Embedded surveillance System Weipeng JING, Xiangfei PAN, Zhenyang LI, Yuelin TANG Journal of Convergence Information Technology, Volume6, Number 8, August 2011

The Application of Embedded Teaching and Experiment Platform: Embedded surveillance System Weipeng JING†, Xiangfei PAN, Zhenyang LI, Yuelin TANG College of Information and Computer Engineering Northeast Forestry University, Harbin, 100044, China doi : 10.4156/jcit.vol6.issue8.39

Abstract Relational Markov networks (RMNs), which extended from Markov random fields (or Markov networks) to the relational setting, are a type of joint probabilistic modeling framework for an entire collection of related entities. RMNs can deal with relational data effectively by integrating information from content attributes of individual entities as well as the links between them. When we learn RMNs, optimization algorithms such as conjugate gradient are combined with approximate probabilistic inference for estimating the parameters. However, the computational complexity of learning an RMN is very high because the approximate inference algorithm must be invoked repeatedly during the whole training process, so the training time is usually unacceptable while the scale of the RMN is too large. In this paper, we propose a two-stage learning framework for RMNs. We divide all of the cliques in an RMN into two categories: evidence cliques and compatibility cliques. In the first stage we estimate the parameters of evidence cliques in the flat setting without take the link attributes into account, and then in the second stage we estimate all of the parameters of both the evidence and compatibility cliques based on the results of the first stage. The experimental results on a bibliographic collective classification task show that the two-stage learning framework can obviously increase the convergence speed of the optimization algorithms and save much training time while learning an RMN.

Keywords: S3C2410, Video Monitoring, GPRS, File Transmission 1. Introduction With the rapid development of modern science and technology (especially microelectronics technology), the embedded technology has been widely used in every field of human society. The human society has entered into on embedded technology as the core of the PC era. With the improvement of people's living standard and safety awareness increasing, a surveillance system can be defined as a technological tool that provides to humans an extended perception and reasoning capability about events of interest[1]. However the traditional image monitoring system mainly is used for complex on-line monitoring in wired [2]. There are little not on-line monitoring products, and the cost of not only cable and optical fiber transmission but also maintenance cost are all high, agile and mobility small general generality is poor, and also the equipment is giant[3]. Therefore it is suitable for application in bazaar, library and other large occasion. And in some occasions, such as shipping important cargos of trucks and unattended precious equipment, it does not need real-time monitoring. Only when the area is invaded, the content is only necessary collected and transmitted scene image and video to server and also send warning signal, in order to achieve the purpose of monitoring and warning. At present several remote monitoring systems based on internet have been proposed [4-5]. In this paper we design based on embedded and GSM/GPRS, video surveillance and alarm system. The pyroelectric infrared module is used to monitor jurisdictions, only if the invasion is detected the video surveillance system will process the image and give alarm. The characteristics of the system not need human operator, especially the cost of data transmission is low. Experimental results show that it is applicable to replace traditional human video surveillance monitor system by the automated one. †

Corresponding author. Email addresses: [email protected] (Weipeng JING).

- 327 -

The Application of Embedded Teaching and Experiment Platform: Embedded surveillance System Weipeng JING, Xiangfei PAN, Zhenyang LI, Yuelin TANG Journal of Convergence Information Technology, Volume6, Number 8, August 2011

2. Hardware Structure In this paper, we designed a prototype system for intelligent automated video surveillance which can be used to replace the human operator to monitor a scene and give alarms. The idea of modular structure is designed in the system which one is control module and other is application functional modules. The unified interface form is used between control module and the functional modules. User can choose different function module according to the different application needs. In the video surveillance system each type data can be simultaneously transmitted without mutual interference. Simultaneously also may according to the market demand, we continue design to expand other function modules. The hardware structure of the embedded monitoring system is shown Fig.1.The real figure of the video surveillance and alarm system is shown in Fig .2.

Figure 1. The hardware structure of the embedded monitoring system

Figure 2. The real figure of the video surveillance and alarm system From Fig.1 we can see the Samsung 2410 is used as development platform; pyroelectric infrared sensor is connected with processor GPIO interface; Flash memory (the U disk) is connected with development board USB port; camera adopts mesh 2000 (built-in OV511 chip) via USB is connected with development board, Camera driving uses Linux built-in ov511 chip driver program; GPRS/GSM module through a serial port is connected; The audio driver is taken 2.6.29 Linux kernel uda1341 driven by platform device driver method, according to the actual development hardware parameters, the device files of uda1341 can be modified and added to the platform. S3C2410A has been developed using the advanced ARM920T core, which has been designed by Advanced RISC Machines, Ltd. The S3C2410 implements MMU, AMBA BUS, and Harvard cache architecture with separate 16KB instruction and 16KB data caches, each with an 8-word line length.

- 328 -

The Application of Embedded Teaching and Experiment Platform: Embedded surveillance System Weipeng JING, Xiangfei PAN, Zhenyang LI, Yuelin TANG Journal of Convergence Information Technology, Volume6, Number 8, August 2011

These resources are enough to satisfy the surveillance system requiring. S3C2410A has 117 multi-functional input/output port pins. Each port can be easily configured by software to meet various system configurations and design requirements. So we can define which function of each pin is used before starting the main program. If a pin is not used for multiplexed functions, the pin can be configured as I/O ports. In this paper the pyroelectric infrared is connected with processor in GPIO interface. Universal Serial Bus (USB) device controller is designed to provide a high performance full speed function controller solution with DMA interface. USB device controller allows bulk transfer with DMA, interrupt transfer and control transfer. In this paper, the camera is connected with development board in USB. The driven of the USB program code is described as following. void UsbMain(void) { U8 fun; rGPHCON = rGPHCON&~(0xf

Suggest Documents