Development Of Monitor Terminal For Digital Audio-Video Network Monitor And Recognition System

1. Introduction Audio-video monitor and identity recognition are applied to various kinds of occasion which requires higher safety. At present, audio and video used in monitor are becoming digitized gradually, and biologic recognition is becoming the major technique in identity recognition. Considering the common of the primitive information that is needed in the audio -video monitor,face recognition and speaker recognition, that is image and voice information, the paper proposes a new method to realize network digital audio-video monitor and recognition. This method is based on network,client / server data processing model. It is especially suitable for solving those problems such as hard to manage, high cost which is based on the independent structure entrance guard management and offered a new solution which is effective, reliable, convenient, and low cost for the modern intellectual entrance guard's system. 2. Main research of the paper 2.1 digital audio-video network monitor and recognition system Presently, entrance guard's system mostly adopts the systematic working way of independence, that means all the entrances and exits of the entrance guard's system are separated. This kind of entrance guard's system exist default as follows: (1) System management is very inconvenient. When personnel alter, collect the user record, the manager of the system must carry the device to the entrances and exits to change and read the data, this is very tedious. In addition, when the entrance guard's system can not work normally, the administrator can't receive the information of damaging as soon as possible, having brought inconvenience to users, even will present the potential safety hazard. (2)The ability to exchange and share is bad. The unattached system doesn't have the ability to share and exchange with the manager in time, and the users are unable to inquire about relevant discrepancy records at any time even more. (3)High costs. The algorithm adopted in the identity recognition technique is more and more complicated, which makes the entrance guard's system more and more complicated, too. At the same time the cost increases sharply. To solve the default listed above, this paper proposes a new system to realize digital audio-video network monitor and recognition. This system combined digital audio-video monitor, face recognition and speaker recognition which belong to biologic recognition technique. It incorporates environment monitor, network transmit and identity recognition into an organic whole and synthesizes digital audio and video, embedded technology, network transmission technology and biologic recognition technology and can be applied to the modern intellectual entrance guard's system. The overall framework of network monitor and recognition system of the digital audio-video is based on Ethernet client/server data processing model. This system is made up of two major parts: monitor terminal that placed on the entrances and exits, control and recognition center which is laid in administrative center of the system. They realize the interconnection through the Ethernet. Using the design philosophy of the thin client, the embedded technology is applied in the monitor terminal. The workflow of network monitor and recognition system of the digital audio and video is as follows: monitor terminal detects the arrival of the people, and sends greets and hint to them, then gathers image and voice of them and sent them to the control and recognition center through network. After receiving the image and voice, the control and recognition center adopts face recognition and speaker recognition algorithm to get the characteristic of people and compares them to the characteristic database that store in center. Then write down this activity in order to administrator inquiry. After that it sends back the recognition result to the monitor terminal to perform the corresponding works. In addition, system can also realize some management conveniently. For example, control and recognition center can check the working state of each monitor terminal and setting up audio communication. The system have a lot of advantages such as convenient to manage, an overall monitor information, a strong ability of mutual and sharing, low costs, easy to expand. 2.2 The design and realization of the monitor terminal The paper adopts the top-down design method in the embedded system. According to the monitor terminal's acquirement in the network monitor and recognition system, the hardware of platform consists of image capture, audio sampling and playing, network interface, switch signal output, embedded processor which operate system and application software. Considering synthetically systematic performance and cost, the paper adopts the kernel of the embedded processor S3C44B0x based on ARM7TDMI core, adopts...