An Autonomous Institution

Affiliated to VTU, Belagavi,

Approved by AICTE, New Delhi,
Recognised by UGC with 2(f) & 12 ( B)
Accredited by NBA & NAAC

Project Phase-1(0th Review)

Title: Finding Missing Person Using AI

Presented By
Chaithanya K M(1MJ20IS020)
Durga Prasad S R(1MJ20IS025)
Meghana R(1MJ20IS049)
Nayana S L(1MJ20IS059)

1
 Contents

1. ABSTRACT
2. EXISTING METHODS
3. PROPOSED SOLUTION
4. APPLICATIONS
5. CONCLUSION

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 2
 ABSTRACT
 Importance of face recognition systems has sped up in the last few decades and it is a part of biometric processing.
 Everyday hundreds of people go missing throughout the world, these people are either kidnapped, taken as slaves or
forced to work underage, or beg on railway platforms, small shops or compelled into prostitution, human-trafficking
or various other illegal activities.
 The posts of missing people are found in our day to day lives on social media platforms, news channels, newspapers
etc. This research work has been proposed to help them locate missing people in an easier way.
 The proposed tool, tends to reduce the time taken to trace the person and improve the process of finding the missing
person and this tool has an Android application for the common citizens as well as a desktop software built using
Python for the police stations by also including a face recognition algorithm based on CNN.
 Choosing the CNN model for face recognition, VGG-Face and proper training of it results in a deep learning model
invariant to noise, illumination, contrast, occlusion, image pose and age of the child and it outperforms earlier
methods in face recognition based missing child identification.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 3
EXISTING METHOD

 Using a Mobile Based Web Service to Search for Missing People.

 Missing Child Identification Using Face Recognition System .

 Missing Person Detection System in IoT .

 Android Based Application - Missing Person Finder .

 Efficient Face Recognition System for Identifying Lost People .

 Drawbacks of Existing System.

 SWF-SIFT is computationally heavy and therefore costs lots of time.

 It is based on Histogram of Gradients where each pixel in the patch needs to be computed.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 4
 PROPOSED SOLUTION
An android app for the common citizens and a desktop software for the police station which has an enhanced facial
recognition algorithm based on CNN.

The modules are:

 Case Registering: - Initially, the civilians capture the photo of suspicious looking people who are working in
places where they should not be. This procedure is quite concise as the details are filled (if available). Also,
the families whose members are lost under certain circumstances can avail this feature.
 Encoding the facial points: - The 68 pair of facial points are recognized are encoded in a cryptographic
format. The encoded text generated is used for further processing. As mentioned, we have the mechanism of
encoding the facial points.
 Traversing through police database: - Once the encoded key is generated, now it is matched with the keys
of other photos This matching of keys can also be understood as matching of faces or facial attributes. Once
a match is found; it is displayed else it shows as a new record in the repository.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 5
Software Tools Used
 Firebase: - The Firebase Realtime Database is a cloud-hosted NoSQL database that lets you store and
sync data between your users in real-time. In this we are store the details of the missing person like his
name, DOB, missing date and time and finally the image of missing person, so that it would be compared
with images that the user uploads.
 Spyder IDE(Python): - Spyder is an open-source cross-platform integrated development environment
(IDE) for scientific programming in the Python language. As python has many in-built functions for
extracting facial points of an uploaded image.

Hardware Requirements
 Mobile or PC with good camera quality: - A camera is required to capture image a person so it is
mandatory to have a mobile or PC with good camera quality.
 Ram capacity: RAM, or Random-Access Memory, is temporary storage for a computer to handle data. A
system should have at least a RAM capacity with 2GB.
 Operating System: An operating system (OS) is system software that manages computer hardware,
software resources, and provides common services for computer programs. Windows and Android and
are the most commonly used OS , so we are designing a application which works on both of these
platforms.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 6
Problem Statement
 “To Identify Missing Person, we are Building an Android Based Application Using CNN
Algorithm”
 Input: Data of the missing person with his photo uploaded and the image of some random
person.
 Output: Display the details of person, if the image is matched with the existing data.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 7
 APPLICATION
 Finding the Missing Person even after a long time is possible.

 The time spent by the police and the people related to the person is hugely saved, thereby,
increasing productivity and efficiency.
 This project can be mainly used by the police officer where a missing persons cases are
registered. With this project the police officer will be able to store the details of missing
person along with his photo. So that in future there a might be chances that some random
person might upload the photo of this lost person so that we can handover him to their
respective guardian.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 8
 CONCLUSION

 Throughout the world, the number of missing people keeps on increasing every day and more than half of
them remain untraced. They are either left to beg or handicapped in such a way that they become almost
unrecognizable.
 Using the proposed system, it is very much possible for the common civilians to interfere in these illegal
activities and try to curb the crimes. A missing person identification system is proposed, which combines
the powerful CNN based deep learning approach for feature extraction and support vector machine
classifier for classification of different child categories.
 This system is evaluated with the deep learning model which is trained with feature representations of
children faces. By discarding the SoftMax of the VGG-Face model and extracting CNN image features to train
a multi class CNN.

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 9
MVJ College of Engineering
Near ITPB, Whitefield
Bangalore-560 067 Thank You

An Autonomous Institute , Affiliated to VTU, Belagavi, Approved By AICTE, New Delhi, Recognized by UGC with 2(f) & 12(B) status, Accredited By NBA and NAAC 10