Design of access control system based on PC/104 platform and face recognition

Abstract: This paper introduces the development trend of access control system, proposes a design scheme of access control system based on PC104/PC104+ embedded development platform and face recognition technology, and introduces the hardware structure and software design scheme of the system.

1. Introduction

In recent years, with the continuous development of science and technology, the access control system as a new type of modern security management technology has been rapidly developed and improved. From the initial password type to today's card type and biometric recognition type, the function and security of the access control system have been greatly improved. At present, the access control system is developing in the direction of digitalization, networking and intelligence.

Common access control systems are mainly composed of three parts: information collection unit, transmission unit, and information identification unit. According to the information communication and processing methods, they can be divided into stand-alone access control systems and network access control systems. Among them, the information processing and collection units of the stand-alone type are combined together, and there is no need for long-distance information transmission between the two. They are often used in small systems or systems with relatively concentrated installation locations; the information processing and collection units of the network type are often not together, and information transmission is achieved between the two through optical fiber, twisted pair, etc., which are often used in large systems or occasions with relatively scattered installation locations. The former has a small investment and is easy to install, but the information processing function is simple, the security is low, and remote monitoring cannot be achieved. Although there are some stand-alone access control systems that can perform biometric identification, their background often requires a microcomputer to run all year round, and the reliability is low; the latter has high technical content, can perform biometric identification, has high security, and is easy to achieve network control and remote monitoring, but due to the influence of network performance, it is not easy to apply to occasions where people frequently enter and exit.

In view of the shortcomings of the above two access control systems, the author proposed a unit access control system design based on PC/104 embedded platform and face recognition technology, which can further realize a distributed access control system integrating distributed identification and centralized management. It not only improves the information processing capability and security of the stand-alone access control system, but also overcomes the problem of paralysis of the network access control system caused by network reasons.

2. System structure design

1. Intelligent community access control system structure

According to the scale of residential area and user needs, the structural block diagram of the access control system designed by the author is shown in Figure 1.

As shown in Figure 1, the access control system of the entire community is mainly composed of the access control system of each unit and the access control management system; the access control system of each unit is composed of the main controller, image acquisition, electric lock, indoor extension, and the access control system of each unit door is connected to the access control management system through the network. Under this structure, each unit door works under the control of the main controller, overcoming the control delay problem caused by network lag and improving the real-time control of the access control system.

2. Access control unit hardware structure

Considering the requirements for the small size and convenient development of the main controller, the PC/104 embedded system is used as the development platform of the system main controller. PC/104, as a "stack PC", has been proven to be a rugged, reliable and cost-effective embedded system solution that can be widely used in commercial, industrial and military fields. Mature desktop PC technology and rich PC software resources make the PC/104 embedded platform one of the most ideal choices for embedded development. As a system compatible with traditional PCs, PC/104 has the characteristics of small size, good openness, high reliability, low power consumption, flexibility and convenience.

The PC/104 embedded module series provides system designers with a set of low-cost, high-reliability, structured modules that can be quickly configured into products. The block diagram of the system main controller built using these modules is shown in Figure 2.

Considering the compatibility between the modules of the main controller, all the main functional modules in the design use PC/104 module boards produced by RTD, including:

(1) CPU module board. The CME47786ER CPU module board is selected. The module has a clock frequency of up to 650MHz, 256MB surface mount memory, 2 RS232 ports, 2 USB ports, 1 10/100Base-T Ethernet port, EIDE controller, supports TFT display, and has a maximum power consumption of 9W.

(2) Video capture module board. The VFG7350ER video capture board is selected. This module supports dual-channel simultaneous capture and has MPEG hardware compression. The dual channels enable the system to capture images from two image acquisition devices at the same time, expanding the system's applicable population, making it suitable not only for adults but also for children with shorter heights.

(3) Data module board. The DM7530HR-8 data module board is selected. This module supports 8-channel differential analog signal input, 2-channel analog signal output, and 16-channel digital I/O with cache. The main controller can control the opening and closing of the electric lock through this module.

(4) Power module board. HPWR104HR-75W power module board is selected. This module can provide a maximum output power of 75W. This module can provide a stable power supply guarantee for the rest of the system modules. For the rest of the modules, some mature products can be selected according to the needs. For example, the storage module can choose the IC25N030ATCS04 6315mm EIDE hard disk produced by IBM, which is used to store the developed programs and the collected user entry and exit images. The keyboard can be designed by yourself according to the needs.

3. System software process design

Based on the above hardware, the unit access control system designed by the author mainly completes the following functions: the information of each user in the unit is stored by household (or department), and a mathematical model of the user's face is generated; after receiving a user's call request, the system realizes image acquisition, automatic face image detection, and recognizes the collected visitor's face image according to the call number, and executes the door opening action after confirming the user's identity, or prompts the indoor personnel to open the door after receiving the door opening instruction from the indoor personnel, otherwise it enters a waiting loop, and if the waiting time is exceeded, the call ends. To realize this function, the software flow is formulated as shown in Figure 3.

As shown in Figure 3, after receiving a user call, the system collects image information and performs face detection. If there is a face, face recognition is performed, otherwise the user is prompted by voice; if the user passes the identity verification, the door is opened, otherwise, the indoor personnel are prompted by voice or sound and wait for the indoor personnel to respond. If the waiting time exceeds, the call ends; after the main controller issues a door opening command, the electric lock opens, waits for the user to pass, and then closes the door, and turns on the tailgating detection device according to the security level. If there is tailgating, a voice alarm signal is issued and the door is closed.

Among them, face detection and recognition are the core of the software system. In this software, a method based on Gabor features and neural networks is used to detect faces; then a method based on principal component analysis and support vector machine is used to compare and identify the collected and detected face images with the images in the reviewed user face library.

4. System Features

From the above analysis, we can see that the unit access control system developed using the PC/104 embedded platform has the following characteristics:

1. Small and flexible: Since the unit access control system designed by the author is developed on the PC/104 embedded platform, this stack-type PC itself is very small. The standard size of the board defined by the PC/104 specification is only 96mm (length) × 90mm (width), which is easy to install in a small space.

2. High reliability: PC/104 boards are extremely reliable in electrical and mechanical characteristics, with low power consumption, less heat dissipation, and good shock resistance. If the environment is more severe, you can also consider using the PC/104 reinforced system IDAN. The aluminum alloy structure of the system not only seals the system, but also plays a role in heat dissipation, which can ensure the normal operation of the system between -40~85℃.

3. Good scalability: Since the module boards used in this design can support both PC/104 bus and PC/104+ bus at the same time, they have good scalability. For example, GPRS communication modules can be added to them to increase remote control functions for the system.

4. Easy to maintain: The designed system is compatible with ordinary PCs. Using some existing desktop PC system software and development software, the designed face detection and recognition software can be easily upgraded, debugged and modified.

V. Conclusion

Aiming at the problems of large space occupation, high maintenance cost and low stability in some unit access control systems, a design scheme of access control system based on PC/104 embedded platform and face recognition technology is proposed. Compared with some existing access control system development schemes based on DSP or single-chip microcomputer, this design scheme has the advantages of short development cycle, stable performance, high security and good scalability, and has broad application prospects.