Criminal Face Recognition Using Raspberry Pi

International Research Journal of Innovations in Engineering and Technology (IRJIET)
ISSN (online): 2581-3048

Volume 3, Issue 11, pp 1-20, Nov-2019
Criminal Face Recognition Using Raspberry Pi

1
Arati Dandawate, 2Vivekanand Vinod Kalaskar, 3Amit Shivappa Vajrashetti, 4Swanand Dnyandev Zarekar,
5
Ramij Shaukatali Shaikh
1
Asst. Prof. Dr. Arati Dandawate, PHD in Image Processing, Asst. Prof. Jaywantrao Sawant College of Engineering, Pune,
Maharashtra, India
2
Vivekanand Vinod Kalaskar, Student, B. E. Computer Engineering, Jaywantrao Sawant College of Engineering, Pune,
Maharashtra, India
3
Amit Shivappa Vajrashetti, Student, B. E. Computer Engineering, Jaywantrao Sawant College of Engineering, Pune,
Maharashtra, India
4
Swanand Dnyandev Zarekar, Student, B. E. Computer Engineering, Jaywantrao Sawant College of Engineering, Pune,
Maharashtra, India
5
Ramij Shaukatali Shaikh, Student, B. E. Computer Engineering, Jaywantrao Sawant College of Engineering, Pune, Maharashtra,
India
Abstract - This paper presents a real time face recognition automatically detecting the human’s face from the databases
using an automated surveillance camera. The proposed this system can be used. In recent years open computer vision
system consists of 4 steps, including (1) training of real has been widely used in different kinds of applications such as
time images (2) face detection using Haar-classifier (3) surveillance camera, robotics etc. This technology is used for
comparison of trained real time images with images from authentication, validation, authorization, and identification. In
the surveillance camera (4) result based on the developed countries, the government creates a datasets which
comparison. An important application of interest is is helpful for recognize the human face which compares the
automated surveillance, where the objective is to recognize suspicious act with trained dataset and information stored in
people who are on a watch list. The aspiration of this database.
paper is to compare an image with several images which
has been already trained. In this paper, this system Face identification [1] is defined in three steps (1)
represents a methodology for face detection robustly in face detection (2) feature extraction (3) face recognition.
real time environment. Haar cascading is one of the Camera configuration is very important to track moving
algorithms for face detection. Here system uses Haar like persons and recognize [2] them precisely. Facial feature points
classifiers to track faces on OpenCV platform. The encode critical information about face shape. Precise location
accuracy of the face recognition is very high. The system and facial feature points tracing are important. Each feature
can successfully recognize more than one face which is point is usually detected and traced by performing a local
useful for quickly searching suspected persons as the search for the better matching position [3]. There are very less
computation time is very low. In India, we have a system researches on face recognition using edge-based detection [4].
for recognizing citizen called Aadhaar. If system uses this The edges are not only carrying valuable data about face but
as a citizenship database, it can differentiate between are also simple to process. The Viola Jones method builds a
citizen and foreigner and further investigate whether the classifier by selecting a few significant features using
identified person is criminal or not. AdaBoost. Viola jones method successfully merges more
composite classifiers in cascade structure [5] which
Keywords: Raspberry Pi, Surveillance Camera, Face exponentially increases speed of detector by focusing on the
Recognition, Image Masking, OpenCV. favourable features of the face.
I. INTRODUCTION II. METHODOLOGY
Face recognition which is a combination of machine This paper is going to develop face detection and
learning and the biometric techniques which holds the recognition system that is capable of processing images very
qualities of not only high precision but also the reliability. For fast while acquiring very high true positive face detection
rates. Object recognition frameworks have been tried
© 2019 IRJIET All Rights Reserved www.irjiet.com Impact Factor: 2.2 1


crosswise over different standard face databases, with and he was under the surveillance of the camera is noted. If he is
without noise and other obscuring effects. The outcome of not a citizen, it is further compared with the images stored in
Object recognition frameworks uncover that well utilized face international watch list database. If match found, the time for
recognition even from low quality pictures and shows which he was under the surveillance of the camera is noted. If
astounding execution productivity, monitor threats, and match is not found in both the watch lists, he is considered as
avoid/examine criminal activities addition to face detection an innocent.
motion detection is always an important requirement.
Criminal Identification Using Raspberry Pi is mainly

about maintaining a secure environment. OpenCV (open III. RESULTS AND DISCRIPTIONS
source computer vision) is the major software that is being
used in this project. For detecting faces, system using various
This section pays special attention to the main
algorithms like Haar cascade, linear SVM, deep neural
consequences of face recognition for criminal identification.
network, etc. The main method that is proposed in this work
This deals with outline development and testing the Face
is, if any person comes in front of the pi camera, first it will
recognition. Here, software’s out there that resemble the same
look for potential matches that we have already stored in this
services as the face recognition system, however, this also has
system. If the module finds a match then it captures the person
its own merits that can contribute to the society.
and sends an alert to the Headquarter.
Due to its intricacy and revelation building, the

recognition of the automated face identification tool might be
burdensome. Moreover, there is also possibility for giving the
interface as user – friendly which wants the minimal amount
of interaction among the users. With this system, which
automatically processes the image, users need only requires to
offer input a real time image & the system will do the rest.
Furthermore, new functions can be added by experienced
developer because this is an open source system. This is a
simple model which always makes the system easier for
utilize. As completely, there are some of the advantages which
have to be recognized as below:
Figure 1: System Architecture
It is a healthier substitute which used for identifying
In this paper, system will use three databases. The the thumb print. Criminal photo captured through a video
images in all the three databases are pre-processed to remove source is fed the identification of the system. This system has
noise and redundancy. Then it undergoes feature extraction the capacity for automatically doing the process like
where Haar cascade is used. First database will have the recognize, detect and extract the features of the images and
images and details of all the citizens of the given country. identifies the actions.
Second will have the images and details of criminals/suspects
who belong to that country. Third database will contain the
images and details of the criminals and suspects who are not IV. CONCLUSION
the citizens of the country. The video obtained from the
This paper has introduced the face recognition system
surveillance camera will be converted into frames. When a
based on Haar cascade. Even though the stored set of images
face is detected in a frame, it is pre-processed where noise and
of the person in the database differ from the input image, the
redundancies are reduced. Then it undergoes feature extraction
system serves as a fruitful method of identifying the faces. In
where Haar cascade comes into picture.
order to find the difference between input image and stored
The processed real time image is compared with the images the system takes out and calculates main features from
processed images already stored in the citizen database. If the input image.
match is found, it is further compared with the images stored
Thus, some changes in the new face image to be
in local watch list database to identify if the person is
recognized can be allowed. Recognition precision and better
criminal/suspect. If he is criminal/suspect the time for which
discriminatory power, computational cost because smaller


images with only main features require least processing to

train the Haar wavelets. The main advantage is that system [9] G. Bradski& A. Kaebler (2009), “Learning OpenCV”,
uses citizenship database which already exists. China: Southeast Univ. Press.
[10] Open Source Computer Vision Library Reference

REFERENCES Manual-intel
[1] International Conference on Intelligent Computing and [11] Gary Bradski& Adrian KaehlerO‟Reilly (2008),
Control, Volume 1, Karpagam College of Engineering, “Learning OpenCV”, O’REILLY Media.
Coimbatore India. June-2017. (Published on IEEE Xplore
Digital library).
AUTHOR’S BIOGRAPHIES
[2] Published in Inderscience (International publishers of
academic, scientific and professional indexed
by Scopus(Elsevier), ACM Digital Library, Asian Digital Vivekanand Vinod
Library, Academic OneFile(Gale)). Volume 8 No. 4 pp. 403- Kalaskar
417 http://dx.doi.org/10.1504/IJCVR.2018.093968
BE (Computer
[3] Jyoti Nautiyal, Shivali Gahlot and Pawan Kumar Mishra, Engineering)
“An automated technique for criminal face identification using
biometric approach”,Conference on Advances in
Communication and Control Systems 2013(CAC2S 2013) Amit Shivappa
Vajrashetti
[4] Unsang Park, Hyun-Cheol Choi, Anil. K Jain and Seong-
Whan Lee, “Face tracking and recognition at a distance: A BE (Computer
coaxial and concentric PTZ camera system”, IEEE Engineering)
transactions on information forensics and security, vol.8,
no.10, October 2013
[5] Yongqiang Li, ShangFei Wang, Yongping Zhao and Qiang Swanand Dnyandev
Ji, “Simultaneous facial feature tracking and facial expression Zarekar
recognition”, IEEE Transactions on image processing, vol.22,
no.7, July 2013. BE (Computer
Engineering)
[6] Reza Moradi Rad, Abdolrahaman Attar, Reza Ebrahimi
atani, “A Robust Face Recognition method using edge based
features”, 2012 IEEE Symposium on Computers and Ramij Shaukatali
Informatics.
Shaikh
[7] Paul Viola, Michael Jones, “rapid Object Detection using a
Boosted Cascade of Simple Features”.
BE (Computer
Engineering)
[8]Face Recognition using Local Binary Patterns (LBP) By
Md. Abdur Rahim, Md. NajmulHossain, Tanzillah Wahid &
Md. ShafiulAzam
*******

Criminal Face Recognition Using Raspberry Pi

Uploaded by

Criminal Face Recognition Using Raspberry Pi

Uploaded by

International Research Journal of Innovations in Engineering and Technology (IRJIET)

ISSN (online): 2581-3048