Development of an Intelligent System for Bank Security

Neeraj Khera Asst. Professor

Deptt. Of ECE, ASET, Amity University Noida, U.P.

nkhera@amity.edu

Amit Verma M. Tech. Scholar

Deptt. Of ECE, ASET, Amity University Noida, U.P

amit8012kruti@gmail.com

Abstract— This paper proposed an effective monitoring and controlling system for bank locker rooms which is completely autonomous. The security system is designed to detect the illegal entrance in the bank locker room areas that commonly happens in cases of the robberies. The major concern with current manually supervised security system is that if the robbery occurs then the banks are not been able to identify the robbers due to lack of proof. The system will focus on the safety of the bank locker rooms in an effective way by detecting and controlling unauthorized motion. The proposed security system will save the images whenever the motion will be detected that can be used in future for investigation The system will communicate the image data continuously to the remote location control rooms using web based monitoring through local area network (LAN) and can also send the warning text short message service (SMS) to the operator using GSM technique.

Keywords—motion detection; GSM modem; microcontroller; website

I. INTRODUCTION For a common human being the bank means a place which

represents a top level of security. On a daily basis we are involved in banking transaction. To secure our expensive jewellery, important documents or cash, we use to use bank locker rooms. It has become an important part of our life. To survive in this competitive world and for a continuous growth, the banking industry needs to provide a high degree of security. Because of the public interest every day new branches are opening. The more number of branches required more security.

Current systems and services are becoming more and more autonomous and the banking sector is not too far from it. Video surveillance [1] in moving areas has become a current topic of interest in computer vision technology. You can see all the branches are under the surveillance of CCTV cameras, alarm systems, emergency buttons etc. The CCTV cameras are used to monitor the unauthorized activity. It needs to be monitored continuously by a human being which is very difficult work; especially in nights. The alarm emergency button also needs to be pressed manually. This conventional system requires lot of man power.

A system can be developed which will automatic detect unauthorized motion and inform to the security officials of the banks by different ways without any need of a human being.

The Microcontroller Based Bank Security System fulfils all these requirements. A prototype of this security system has been designed in the dissertation to increase the level of security in bank locker rooms effectively. The motion detection will be done through camera [2],[3],[4] itself and the hardware associated with it will provide 3 different ways to inform the security officials i.e. using alarm system [5], a warning message and the image which has detected the motion [6],[7] will be automatically uploaded on webpage which can be downloaded from anywhere. For messaging a GSM module [8],[9],[10] will be utilized.

So the important objectives of bank security system are tracking the bank locker room areas, detection of motion and taking the necessary control action. The further sections will describe that how these objectives have been achieved.

II. SYSTEM BLOCK DIAGRAM The following section will tell us about the approach to the

design to achieve the required goals. The design of the security system has been shown in figure 1.

The figure shows that the system design comprises various hardware modules. The primary modules are camera for visual surveillance, CPU for motion detection and to upload the motion detected image into the website. The CPU is also connected to the microcontroller via serial port. The system requires microcontroller with two serial ports; one for the CPU and another for the GSM modem. To fulfil the requirement of two different serial ports, we are using two different microcontrollers because Microcontroller ATMEGA16 is provided with only one serial port. The microcontrollers are connected to each other through port C.1. The microcontroller 2 is connected to the alarming system to activate it and GSM modem to send the warning message. All the hardwares are properly interconnected with each other and well insulated to avoid short circuit issues.

III. SYSTEM IMPLEMENTATION This section will illustrate about the different

methodologies opted to implement the security system and the workflow of the system. The section will also illustrate different program codes used for the implementation of the system.

319978-1-4799-4236-7/14/$31.00 c©2014 IEEE

Fig. 1. Block Diagram of Microcontroller Based Bank Security System.

A. Methodology for Motion Detection For motion detection we are using MATLAB software. The

motion has been detected by continuously monitoring the locker room area through camera and to compare the current image and compare it with the reference image. There are various functions available in MATLAB for image matching and comparison. In this security system we are using imabsdiff command to identify the motion. The syntax for this command is

(1)

where, X and Y are array of integers which gets compared by the function and an absolute difference will be calculated which will stored in the output i.e. Z. The basic function of this command is to perform subtraction between corresponding element of X and Y. For normal numerical data it performs X-Y and for the logical data it performs X xor Y. As we know that MATLAB keeps all the data in matrix form so let’s take an example for better understanding of this function.

The figure 2 shows how the motion will be detected using the imabsdiff function. The top left section shows the reference image X. The top right section shows the current image Y which is going to be compared with the reference image. When the two images are compared and subtracted using the imabsdiff function it will generate a new image Z which is shown in the bottom left section. The figure clearly shows that some motion has occurred because after subtraction the pixels, which have detected the motion, will have values much. When the subtracted image will be converted in black & white image then it will give a very clear picture as shown in bottom right section, that there are number of white pixels in the frame so the motion has been identified.

Fig. 2. Motion Detection through MATLAB

B. Methodology for Website Updatation To upload the image a webpage has been designed in

notepad and it has been saved with .html extension. The html code can be given as:

<echo?>

Welcome to the 3 Tier Bank Security System Web Security Portal

</echo>

<a href="motion.jpg">Download the Motion Detected Image</a>

C. Methodology for Sending Warning Message

The warning message will be send to the security officials using the GSM modem. The GSM modem will be connected to the serial port of microcontroller 2 through MAX 232 IC. Whenever the motion will be detected by the camera the microcontroller 2 will send AT commands to generate a message from the GSM modem. Series of AT commands are required to generate a message. The AT commands to generate a message used in this security system are:

AT

AT+CMGF=1

AT+CMGS=”Destination Phone Number”

Message Body

Ctrl+Z

D. Flowchart of the Security System

Figure 3 shows the basic flowchart of the system. The first step will be to switch on the camera using the camera and to provide the power supply to all the components of the system. Now the camera will capture the reference image using the MATLAB software. Once the reference image will be set the

320 2014 5th International Conference- Confluence The Next Generation Information Technology Summit (Confluence)

detection of motion will start. The current image will be taken continuously and the motion will be checked using the MATLAB software. If the motion will not identify it will again take the current image and check again for the motion. This process will continue untill the motion will identify. Once the motion will identify the motion detected image will be upload into the website ans an activation signal will be sent to microcontroller 1 through serial command. Now the microcontroller 1 will also activate microcontroller 2. The microcontroller 2 activates the alarm system and also sends AT commands to the GSM modem serially so that GSM modem can send warning message to the security officials.

Fig. 3. Flow Chart of the System

IV. RESULTS In figure 4 the output of the MATLAB is shown. The

command window shows that the MATLAB software counts the number of pixels of the current image differs from the reference image and according to the threshold level the motion has been detected or not. In this figure motion detection has been referred has Image mismatched. We have taken count 0 as threshold level in this prototype. Anything more than that can be taken as motion. The threshold level also depends upon the positioning and area of the locker room. To get the better

results it is advice that the threshold level must be set after positioning of the camera and with proper practical observations. The motion detection algorithms are implemented and communicated to remote user as HTML file as shown in figure 4.

Fig. 4. Snapshot of MATLAB Result

Using the standard web browser the communicated HTML page will be open as shown in figure 5. The figure shows that the motion detected image has been uploaded in the website which refers to the motion.jpg file in the html code. The motion.jpg file should be stored in the same hard disk location of host computer where the HTML page has been stored.

Fig. 5. Snapshot of Webpage Result

V. CONCLUSION This paper presents a low cost, intelligent and secured

system for the bank locker room areas. The system will focus

2014 5th International Conference- Confluence The Next Generation Information Technology Summit (Confluence) 321

on the safety of the bank locker rooms in an effective way by detecting and controlling unauthorized motion. The proposed security system will detect the motion through camera and provides security by taking three different control actions. In this system visual surveillance algorithm has been implemented for motion detection on the host computer and the resulting image data has been communicated to client side computers for web based monitoring through local area network (LAN). The unauthorized image detection signal is also online communicated to the microcontroller that sets the warning alarm nearby also send the warning text short message service (SMS) to the operator using GSM technique for taking necessary control actions. The proposed intelligent system has advantages over the current security system like low cost, stable link, large coverage area, and low communication cost.

REFERENCES [1] Robert T. Collins, Alan J. Lipton, Hironobu Fujiyoshi, and Takeo

Kanade, “Algorithms For Cooperative Multisensor Surveillance,” In Proc of the IEEE, vol. 89, no. 10, Oct. 2001, pp. 1456-1477.

[2] Rita Cucchiara, Costantino Grana, Massimo Piccardi and Andrea Prati, Member, “Detecting Moving Objects, Ghosts, and Shadows in Video Streams,” IEEE Transactions on Pattern Analysis And Machine Intelligence, vol. 25, no. 10, Oct. 2003, pp. 1337-1342.

[3] Piyush Mayank and Sudipta Mukhopadhyay, “Temporal Correlation and Probabilistic Prediction Based Face Detection Framework in Real Time Environment,” IEEE Proceedings of 4th International Conference on Intelligent Human Computer Interaction, Kharagpur, India, December 27-29, 2012.

[4] Weiming Hu, Tieniu Tan, Liang Wang, and Steve Maybank, “A Survey on Visual Surveillance of Object Motion and Behaviors,” in Proc. of IEEE Transactions on Systems, MAN, and Cybernetics vol. 34, no. 3, Aug. 2004, pp. 334-351.

[5] Sadeque Reza Khan, “Development of Low Cost Private Office Access Control System(OACS),” International Journal of Embedded Systems and Applications, vol.2, no.2, June 2012.

[6] Ashit Talukder and Larry Matthies, “Real-time Detection of Moving Vehicles using Dense Stereo Objects from Moving and Optical Flow,” in Proc, of IEEE International Conference on Intelligent Robots and Systems, Sep 28 - Oct 2, 2004, pp. 3718-3725.

[7] Ran Eshel and Yael Moses, “Homography Based Multiple Camera Detection and Tracking of People in a Dense Crowd,” 2008.

[8] A. Alheraish, “Design and Implementation of Home Automation System,” IEEE Transactions on Consumer Electronics, vol. 50, no. 4, Nov 2004, pp. 1087-1092.

[9] Genghuang Yang, Boying Wen and Guohua Gao, “Devices for Automatic Irrigation Based on GSM Network and Radio Communication,” in Proc, of International conference on Automation, Robotics and Vision, pp. 2216-2220, 2006.

[10] Ma Yuchun, Huang Yinghong, Zhang Kun and Li Zhuang, “General Application Research on GSM Module,” in Proc, of International Conference on Internet Computing and Information Services, , pp.525-528,2011.

[11] Hesong Huang, Hongning Bian and Shuchuan Zhu, “A Greenhouse Remote Monitoring System Based on GSM,” in Proc, of International Conference on Information Management, Innovation Management and Industrial Engineering, pp. 357-360, 2011.

[12] Nitika and Simpel Jindal, “Multimessaging System Using Gsm Modem,” International Journal of Advanced Engineering Sciences and Technologies ,vol. 11, no. 1, pp. 157 – 161, 2011.

[13] Adamu Murtala Zungeru, Ufaruna Victoria Edu and Ambafi James Garba, “Design and Implementation of a Short Message Service Based

Remote Controller,” International Conference on Computer Engineering and Intelligent Systems, pp. 106-118, 2011.

[14] D.V. Gadre, “Programming and customizing the AVR Microcontroller,” McGraw-Hill, New York, 2007.

.

322 2014 5th International Conference- Confluence The Next Generation Information Technology Summit (Confluence)