© 2015 IJSRSET | Volume 1 | Issue 2 | Print ISSN : 2395-1990 | Online ISSN : 2394-4099

Themed Section: Engineering and Technology

Face Recognition using Edge Detection of Layer

G. Akhila Goud, B. Hemanth
ECE, Hyderabad Institute of Technology and Management, Hyderabad, Telengana, India

ABSTRACT

Face recognition of people in a crowd is a challenging issue that received much attention during recent scenario
attributable to its numerous applications in different fields. The analysis of human crowds has widespread uses from
law enforcement to urban engineering and traffic management. Face recognition in a heavy crowd is one of those
challenging problems and till ages no proper solution has been derived, there is no technique that provides a robust
solution to all. This paper presents a new technique for human face recognition in a crowd in metropolitan cities.
Keywords: Face recognition, discrete cosine transform, Kohonen Rule, Euclidean distances.

I. INTRODUCTION edges found in the current image of the sequence and

The face recognition system is a computer based those present in the background of the scene. The
application for automatic identification or verification of background reference image is derived from the images
a person through a digital image or a video source of the sequence. There are several advantages in
.There are numerous ways to perform this identification creating an image of the background edges for the
or detection. One of the ways is by comparing selected detection of moving objects [Karmann 1990]. The first
facial features from the image and a facial database. This one is that edges are features rather insensitive to
is frequently used in security systems and has a variations of lighting conditions which characterise
flexibility be compared with other biometrics such as outdoor scenes. Another advantage of this approach,
eye iris or fingerprint recognition systems. based on a low level analysis, is that it is independent of
the shapes and the number of moving objects. We
present in this section the principle of the motion
The facial recognition algorithm identifies facial features detection algorithm.
by extracting features, from an image of the subject of
interest. Few algorithms normalise a gallery of face A. Principle of the algorithm
images from a complete set and then compress the face
data, this saves only the data in the image that is useful Let Ik(P) be the value of the grey level function at a pixel
for face recognition. The acquired image is then P in the Kth input image of the sequence for K €
compared with the face data. [1,….,N].A simple gradient operator is applied to each
input image lk .Let Gk denote the resulting current edge
II. METHODS AND MATERIAL image whose magnitude Gk(P) at a pixel P indicates the
likelihood that the pixel P is part of an edge of the
current edge image. This current edge magnitude of the
MOTION DECTECTION PROCESS
image at pixel P. The construction of this reference edge
image is presented in details in the next section. The
The proposed real-time motion detection algorithm is
main idea behind the proposed algorithm is to compare
based on the construction of a structural model of the
the current edge image to this reference edge image
background [Vannoorenberghe 1995]. This model is
which is constructed as to contain only static edges of
composed of all stationary edges in the scene. Moving
the background and capture the layout of the faces and
edges are detected by means of a comparison between

IJSRSET152296 | Received: 31 March 2015 | Accepted: 09 April 2015 | March-April 2015 [(1)2: 378-380]
378
scan with the target layer to match the face so it makes
easy and faster way to scan the crowd. Here we use
the Canny edge detection. It is a unique edge
detection technique that uses a multi-staged algorithm to
detect a wide range of distinct edges in images with
great precision. Figure 3. Selected images

Consider the input to the system is Fig.1 and a person’s

face is to be detected. The Fig.2 is the resultant image
obtained after the application of grey level canny
algorithm. The initial image assumed for the operation is
mostly a image with more than two faces. It can be
exactly two persons in the image out of which one has to Figure 4. Output Obtained
be recognised or in other sense it can also be more than
two, but the minimum number of possible faces required III. RESULTS AND DISCUSSION
is two. The maximum number of faces in the image that
FACE DETECTION BLOCK DIAGRAM
is to be operated depends on the resolution of the image.
It may vary from one resolution to other. Fig.2 is the
The Basic block diagram of the Face Detection and
selected portion of interest. These pictures are matched
recognition system consist of six major blocks in it.
with the database and the resultant out is matched
Prior the major block is the input block. Here the
specifically.
surveillance camera acts as the input accumulator in the
total system. Surveillance means monitoring of activities,
behaviour or other altering information, generally of
people for the purpose of managing, directing, or
safeguarding them. This can include observation from a
distance by using of an electronic device or interception
of electronically transmitted information and it can
include simple, relatively no- or low-technology
operation such as postal interception and human
intelligence agents. The word surveillance was derived
from a French phrase for "watching over”.

Figure 1. Original Image

Figure 2. Crowd edge scan image

Figure 5. Block diagram of Face Detection System

International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)

379
The main heart of this block diagram is Server Database
System. A database system is a collection of information
that is managed so that it can easily be managed,
accessed and updated. In one words, databases can be
classified according to types of information it holds like:
full-text, numeric, and images etc.., There is another
block that acts as the memory element in this operation
that is it stores copies of layers of images. Display unit is
an output device.

IV. CONCLUSION

Thus a system that can detect and identify a person from

a crowd can be designed using a canny edge detection
algorithm. This system is quite efficient when the crowd
size is vast. The crowd image recognition can be used
also in military applications.

V. REFERENCES

[1] Canny, J. F (1983) Finding edges and lines in images, Master's

thesis, MIT. AI Lab. TR-720.
[2] Canny, J. F (1986) “A computational approach to edge
detection”, IEEE Transaction on Pattern Analysis and Machine
Intelligence, 8, 679-714.
[3] Ramadevi, Y & et al (2010) “Segmentation and object
recognition using edge detection techniques”, International
Journal of Computer Science and Information Technology, Vol
2, No.6, 153-161.
[4] Roberts, L (1965) “Machine Perception of 3-D Solids”, Optical
and Electro-optical Information Processing, MIT Press.
[5] Robinson. G (1977) “Edge detection by compass gradient
masks”, Computer graphics and image processing, 6, 492-501.
[6] Rousseeuw, P. J & Leroy, A (1987) Robust Regression and
outlier detection, John Wiley & Sons, New York.
[7] Senthilkumaran. N & R. Rajesh (2009) “Edge Detection
Techniques for Image Segmentation – A Survey of Soft
Computing Approaches”, International Journal of Recent
Trends in Engineering, Vol. 1, No. 2, 250-254.
[8] Sowmya. B & Sheelarani. B (2009) “Colour Image
Segmentation Using Soft Computing Techniques”, International
Journal of Soft Computing Applications, Issue 4, 69-80.

International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)

380