You are on page 1of 18

ADHIYAMAAN COLLEGE OF ENGINEERING (AUTONOMOUS)

HOSUR
DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

MINI PROJECT

SUSPECT TRACKING SYSTEM

BATCH NO: B9 TEAM MEMBERS


GUIDE NAME:
MAGESHKANNAN A (AC16UCS069)
Mrs. E. KANIMOZHIVEENA, M.E
MANOHARAN S (AC16UCS070)
MUKESHKANNA G (AC16UCS079)
ABSTRACT

 In this project Haar cascade algorithm is used for face detection and tracking.
 Haar cascade algorithm is one of the efficient fast face detection and tracking
algorithm.
 This project will be implemented in public cameras located in various places.
 If suspect pass through camera, the camera will detect the suspect and generate a
report.
 With the help of this generated report, we can locate the suspect and the police can easily
track the suspect.
OBJECTIVE

 The objective of the “Suspect tracking system” is to detect and recognize suspect from the
video frames.
 This application detects the suspect very fast.
 The efficiency of the algorithm is high.
 The training time of suspect’s images is very fast.
LITERATURE SURVEY

 FACE RECOGNITION USING HISTOGRAM OF ORIENTED GRADIENTS [Milind


Rane, Sourabh Hanamsheth]
 AN ANALYTIC APPROACH 3D SHAPE DESCRIPTOR FOR FACE RECOGNITION
[Shilpi Singh Tapas Kumar]
 Problem Identified
The drawback of the present system is that it can only linearly traverse to detect the
faces. This process takes much more time to linearly search for the faces so this will not be an
efficient method to detect faces.
SYSTEM SPECIFICATIONS

 HARDWARE SPECIFICATION:
System : Core i5.
Hard Disk : 1TB.
RAM : 4GB.
Monitor type : 15.6’ Color Monitor.
Keyboard type : Standard Keyboard.

 SOFTWARE SPECIFICATION:
Operating System : Windows 10.
IDE : Python IDLE.
Coding Language : Python.
ARCHITECTURAL DESIGN
MODULE 1

 IMAGE TAKING
(1) By using Haar-Cascade method faces are pre trained with more than 6000 sets of
faces.
(2) By using this technique, a human face is identified and cropped in to 192 px X 112 px.
Then that image will be converted into grayscale format.
(3) It automatically stores the hundred images and the name of the image starts from 1 to
60. The numbers were automatically incremented and will be stored as image names.
(4) The name of the person is passed to machine through command line argument and it
creates a folder in the machine and the sixty images will be stored under the given name of the
person.
MODULE 2

 LEARNING MODULE
(1) The data from the face database is taken and it creates a XML file.
(2) The face database contains matrix of the gathered faces and it is created as an XML
file.
(3) XML file contains the average faces of the person’s label and under the label the
matrix of the images are stored.
MODULE 3

 IMAGE TRACKING MODULE:


(1) It follows the Face tracking process from the starting to grayscale conversion.
(2) Then the converted image is cross checked with the faces in the database for any
match.
(3) If any match is found by the machine then it displays the name of that person
otherwise it displays person unknown.
ALGORITHMS / TECHNIQUES:

 HAAR CASCADE
(i) Object Detection using Haar feature-based cascade classifiers is an effective object
detection method. It is a machine learning based approach where a cascade function is trained
from a lot of positive and negative images. It is then used to detect objects in other images.
(ii) First, a classifier is trained with a few hundred sample views of a particular object
called positive examples that are scaled to the same size and negative examples - arbitrary
images of the same size.
(iii) After a classifier is trained, it can be applied to a region of interest in an input image.
The classifier output is “1” if the region is likely to show the object, and “0” otherwise.
SCREENSHOT
IMAGE TAKING
IMAGE SAVED
IMAGE TRAINED
IMAGE TRACKING
GENERATED FILE
REFERENCES:

 GARY BRADSKI AND ADRIAN KAEHLER – Learning OpenCV.


 MARK LUTZ – Programming Python.
 NICHOLAS LOCASCIO AND NIKHIL BUDUMA – Fundamentals of Deep Learning.
 https://en.wikipedia.org/wiki/Haar-like_feature
 https://www.tutorialspoint.com/python_deep_learning
THANK YOU

You might also like