Missing Child Identification System

using Deep Learning and Multiclass

TEAM MEMBERS :
Gudla Lokeswari (19B81A0566) PROJECT GUIDE:
Gullapudi Prasanth (19B81A0567) M. Madhava rao
Assistant professor
Revanth Sai G (19B81A0568)
Naishadha Gurram (19B81A0569) PROJECT COORDINATOR:
Shreya Guthikonda (19B81A0570) Dr. K. N. Madhavi Latha
Associate Professor
 Content
• Abstract
• Problem Statement
• Literature Review
• System Architecture
• Existing System
• Proposed System
• System Configuration
• Flow chart
• Design
• Methods
• Implementation
• Conclusion & Future Scope
 Abstract
• This presents a novel use of deep learning methodology for identifying the reported missing child from the
photos of multitude of children available, with the help of face recognition.
• The public can upload photographs of suspicious child into a common portal with landmarks and remarks. The
photo will be automatically compared with the registered photos of the missing child from the repository.
•Compared with normal deep learning applications, our algorithm uses convolution network only as a high level
feature extractor and the child recognition is done by the trained KNN classifier.
• Classification of the input child image is performed and photo with best match will be selected from the database
of missing children.
• The Convolutional Neural Network (CNN), a highly effective deep learning technique for image based
applications is adopted here for face recognition. Face descriptors are extracted from the images using a pre-trained
CNN model VGG Face deep architecture.
 Problem statement

• As we know that the India is a second largest country in the world if it comes to population.
• As there is a great saying “TODAYS CHILDREN ARE TOMORROWS CITIZENS”.
• Children are being Kidnapped or Missed in crowd places and any religious or social gatherings
etc.
• There are many situations we see a child who is missed in one state may found in another state.
• On an average 174 children go missing every day, half of them remain untraced.
• As per NCRB report which was cited by the MIHA, more than 1lakh children were reported
missing.
 Literature Review
Title of Paper Description Publication Details
“Deep learning", Nature. A deep learning [1] architecture Y. LeCun, Y. Bengio and G.
e considering all these constrain Hinton (2015).
is designed here.

 "Face recognition using Earliest methods for face O. Deniz, G. Bueno, J. Salido
histograms of oriented recognition commonly used and F. D. la Torre,
gradients", Pattern Recognition computer vision features such as (2011).
Letters. HOG.

"Face recognition using sift Earliest methods for face C. Geng and X. Jiang,
feature”. recognition commonly used  IEEE International Conference
computer vision features such as on Image Processing(ICIP).
SIFT.
 Literature Review
Title of Paper Description Publication Detals
"Deep Face Recognition”.  A very deep CNN called VGG-
Face network[8] is used for face O. M. Parkhi, A. Vedaldi and
recognition and its architecture A. Zisserman, (2015).

“Very deep convolutional Convolutional Neural Networks Simonyan Karen and

• If any child is missing, we have to dial 100 for Police or 1098 for Child line.
• Go to the police station and register an FIR.
• We will start searching for the child by enquiring friends and relatives to find clues about
child.
• The child missing from one region may be found in another region or another state, for
various reasons.
Disadvantages:
• So even if a child is found, it is difficult to identify him/her from the reported missing cases.
• It takes more time for investigation.
 Proposed System

•An idea for maintaining a virtual space is proposed such that missing cases are saved in a
repository.
• Here we propose a methodology for missing child identification which combines facial
feature extraction based on deep learning and matching based on KNN.
• The proposed system utilizes face recognition for missing child identification. This is to
help authorities and parents in missing child investigation.
• VGG Face recognition results in deep learning model invariant to noise, illumination,
contrast , image pose and age of the child.
 SYSTEM CONFIGURATION

Hardware requirements:
• Processer : Intel i5
• Ram : 4GB
• Hard Disk : Min 250 GB

Software Requirements:
• Operating System : Any Windows (from version 7)
• Programming language :Python 3.7
• IDE: PyCharm
SYSTEM ARCHITECTURE
FLOW CHART yes no
Unauthorized user
Check

Child details

Preprocess Dataset

Build Resnet50,VGG16 & CNN Model

View Public Upload Missing Child

End process
 User

yes no
Unauthorized user
Check

Public upload Suspected Child

Preprocess Data

Analysis Suspected Child

End process
DESIGN

 UML DIAGRAMS:
A UML diagram is a diagram based on the UML (Unified Modeling
Language) with the purpose of visually representing a system along with its
main actors, roles, actions, artifacts or classes, in order to better understand,
alter, maintain, or document information about the system.
USE CASE DIAGRAM
CLASS DIAGRAM:
SEQUENCE DIAGRAM
COLLABORATION DIAGRAM
ACTIVITY DIAGRAM
METHODS USED IN PROJECT:

 CNN
 VGG
 Resnet
IMPLEMENTATION
CONCLUSION

 A missing child identification system is proposed, which combines the powerful CNN based deep learning
approach for feature extraction.
 By discarding the softmax of the VGG-Face model and extracting CNN image features to train a multi
class SVM, it was possible to achieve superior performance.
 The classification achieved a higher accuracy of 99.41%
 In missing child project student asking to implement RESNET 50 and VGG 16 and compare their accuracy
with CNN
 FUTURE SCOPE OF PROJECT

 In the future, we are getting to extend this technique further by connecting our
system to public cameras and detect faces real-time.

 The frames are going to be continuously sent by the general public cameras to
our system where our system is going to be continually monitoring the frames.

 When a lost person is identified in any of the frames, it'll be notified to the
concerned authorities
Thank You