Description

ABSTRACT

Computer Vision and Internet of thing (IOT) is the emerging technologies now a days. To reduce the man power for security aspects in this project a Surveillance system has been implemented using a single board computer i.e. Raspberry Pi 3 which will act as the CPU in which we will do the coding part using Python and a module named Open Source Computer Vision (OpenCV). A local database of the authorized person is been made which has the images of all the persons who are authorized to enter that security area. Camera will act as the computer vision which is in surveillance mode, which will automatically take the image of the person compare with the local database if the person matches it will open the door and if the person in unauthorized then it will send the SMS and the image of the unauthorized person through MAIL using the SMTP. Also a webpage is made which will show the unauthorized person image. This image will be stored in the apache server where we have our web page.

 

INTRODUCTION

In this project OpenCV module of python plays a key role to recognize the person if the person the authorized or unauthorized. If the person is authorised it will open the door (prototype with DC Motor). If the person is unauthorised it will capture his image and send to the mail using SMTP. Also hardware like Buzzer and dc motor is being embedded which represents the alarm and door in this project.

Existing system

In the existing system, PIR sensor is used to recognize if the person is there or not but it will not click the perfect image of the person as the person may be facing the camera or may not be facing the camera. Accuracy of the person will differ.

 

 Proposed system

In this proposed system, as the camera is in surveillance mode automatically the person image will be captured as the person comes in front of the camera, it will compare with the database using the convolution neural networks, where the accuracy will be very high  and the speed also will be high.

BLOCK DIAGRAM

BLOCK DIAGRAM DESCRIPTION

• In this project camera is used which will be in the surveillance mode.
• A database is made of the authorized persons.
• Connect USB camera with raspberry pi
• Connect power supply for Raspberry pi
• Plug the HDMI cable in Raspberry pi from the monitor using VGA to HDMI converter cable
• Connect USB Mouse and USB keyboard to the Raspberry pi
• As the person is detected image is captured and is being compared with the database.
• If it matched door will open if it doesn’t then a mail with the person image attached will be sent.

 

 HARDWARE REQUIREMENTS

• Raspberry Pi
• USB Camera
• DC Motor
• Buzzer
• Relay circuit
• SD card
• Monitor

 SOFTWARE REQUIREMENTS

• Raspbian Jessie
• HTML and PHP
• Python with OpenCV
• Language – Linux

RESULT

 

• Above figure shows the output on the monitoring screen, when the person come in front of the camera it will automatically detect the person.
• Captures the image and start compariring with the database.

• Above figure shows the output in the monitoring screen after training i.e. when the person is detected it will compare with the database.
• As the above image displays that the person is unauthorized it will send the mail to the person image using SMTP.

 

• If the person is authenticated it will display the person name and also the status i.e. the person is authenticated or not as shown in the above image.

 

CONCLUSION

According to this system, it make our life easy and more secure as we can get the status of the person entering into the house from a remote area with the image of the unauthorized person through SMTP mail protocol and also a SMS API is made using TWILIO which will act as indicating that a unauthenticated person is entered into the room.

 

REFERENCES

[1] “An improved face recognition method using local binary pattern method” in 2017 11 th International Conference on Intelligent Systems and Control (ISCO)

[2] “An advancement towards efficient Face Recognition using live video feed” in 2017 International Conference on Computational Intelligence and Networks.

[3] M. Li and B. Yuan, “2D-LDA: A statistical linear discriminant analysis for image matrix”, Pattern Recognition Letters, vol. 26, pp.527- 532, 2005.