FORENSIC FACE SKETCH
Construction
And Recognition
AGENDA
01 INTRODUCTION
04 MODULES IMPLEMENTED
02 PROBLEM STATEMENT
05 ADVANCEMENT
03 EXISTING SYSTEM &
PROPOSED SYSTEM 06 TECHNOLOGIES
IMPLEMENTED
INTRODUCTIO
N
"Forensic FaceSketch: Revolutionizing facial ID with advanced machine learning.
Real-time sketches, dynamic recognition, ethical integration for precise criminal
investigations."
PROBLEM
STATEMENT
Traditionally, identifying criminals with hand-drawn sketches is slow and limited.
Leveraging technology for real-time matching is crucial. An app for rapid and
accurate facial sketch creation streamlines crime-solving.
EXISTING AND PROPOSED
EXISTING: PROPOSED:
Utilizing deep learning, FACES generates The proposed system involves developing
facial sketches from descriptions and software that constructs facial sketches based
employs facial recognition for suspect on descriptions and performs face recognition.
identification.
DESIGN AND
IMPLEMENTATION
NO SELECT FACE
LOGIN ROUGH ROUGH SKETCH
SKETCH FEATURE
RENDER IMAGE
YES
(IMPORT DRAWN SKETCH)
FEATURES EXPORT SKETCH
EXTRACTION
CRIMINAL FACE
DATABASE
RECOGNITION
OUTPUT
Authorized user CNN Model ML Model Database Login Application User
Provide Credentials
Validate Credentials
Provide Description
Create Face Sketch
Extract Features(Sketch)
Train Model(Sketch)
Provide Photo
Extract Features(Photo)
Analyze Features(Sketch,Photo)
Recognize Face
Display Output
Authorized user CNN Model ML Model Database Login Application User
MODULES
IMPLEMENTED
1 2 3 4 5 6
LOGIN DATABASE CONSTRUCTION OF COLLECTION LIST FEATURES EXTRACTION FACE RECOGNITION
SKETCH
LOGIN
The system prioritizes privacy and security from the login stage. It first verifies the
host machine's identity using unique identifiers, then employs Two-Step Verification,
sending real-time OTPs to registered emails for added security.
LOGIN MODULE
CLASSES
METHODS
LIBRARIES
Login_screenController • Random()
• Sendotp()
• javax.mail.*; • Login()
• java.util. • initialize()
• java.net.*; • handleButtonAction()
• java.sql.*;
• java.util.logging.*;
• javafx.fxml.*;
• Google Authenticator Library (java-
libpam):
• Apache Commons Codec
DATABASE
We use Amazon S3 to store and retrieve any amount of data at any time, from anywhere.
Amazon S3 stores data as objects within buckets. An object consists of a file and optionally
any metadata that describes that file.
DATABASE MODULES
Library Classes Methods
• java.sql.Connection; connectdb
ConnectDB
• java.sql.DriverManager;
• java.sql.SQLException;
CONSTRUCTION OF
SKETCH
Facial sketch construction in forensic art involves interpreting verbal descriptions,
sketching facial features with anatomical accuracy, seeking feedback, and finalizing
realistic drawings for investigative purposes.
CREATE A NEW FACE SELECT THE FACE
SKETCH ELEMENT
SELECT A FEATURE TYPE
FROM THE DATABASE
MATCHES THE NO SELECT A HEAD
EYEWITNESS TYPE
DESCRIPTION
YES
SELECT ANOTHER SKETCH YES SAVE THE FACE RECOGNITION
MODULE
COMPLETED?
FACE ELEMENT SKETCH
FACE SKETCH NO
CONSTRUCTION
CONSTRUCTION MODULE
LIBRARY CLASSES METHODS
• javafx.scene.image.*;
• javafx.fxml.FXML.*; • DashboardController • initialize
• javafx.embed.swing.SwingFXUtils.* • drag
• javafx.scene.*; • dragSketch
• javafx.stage.*; • save_img
• javax.imageio.ImageIO; • toolKit
• java.io.*; • onSave
• java.util.*; • onCompare
• onReset
• onHeadSelect
• onHairSelect
• onEyesSelect
• onEyeBSelect
• onNoseSelect
• onLipsSelect
• onMustSelect
• onMoreSelect
COLLECTION LIST
Collection list fetches faces stored in the specified collection using the Amazon
Rekognition service and prints information about each face as JSON strings. It
handles pagination to retrieve all faces in the collection. This could be useful for
tasks such as monitoring the faces stored in a collection, analyzing facial
recognition results, or managing face data within an application.
COLLECTION LIST MODULES
Library Classes Methods
• ListFacesRequest()
• java.util.list; connection_list
• withCollectionId(collecti
• com.amazonaws.services.reko
gnition.AmazonRekognition onId)
• com.amazonaws.services.reko • listFacesResult.getFaces(
gnition.model.Face; )
• com.fasterxml.jackson.databin • rekognitionClient.listFac
d.ObjectMapper es(listFacesRequest)
• objectMapper.writerWith
DefaultPrettyPrinter().wr
iteValueAsString(face)
FEATURES
EXTRACTION
Feature extraction involves identifying key aspects of facial sketches, such as
shape, proportions, and unique traits, using algorithms to enable accurate matching
and recognition in forensic investigations.
FEATURES EXTRACTION
MODULES
Library Classes Methods
• com.amazonaws.services.reko • collection_search_face
• recognition
gnition.*;
• com.fasterxml.jackson.databin
d.ObjectMapper;
• java.util.List;
• javax.swing.JOptionPane;
FACE RECOGNITION
Face recognition is the process of identifying or verifying individuals from facial
images. It involves analyzing facial features, patterns, and structures using
algorithms for authentication or identification purposes.
FACE SKETCH RECOGNITION
MODULE
HAND
DRAWN
OPEN FACE SKETCH TYPE SELECT THE
RECOGNITION
SKETCH
MODULE
USING THE
APPLICATION
SKETCH
UPLOADED TO
SERVER
MATCH THE
SKETCH WITH
DATATBASE
DISPLAY OTHER
DISPLAY SHARE THE
DETAILS ABOUT
SIMILARITY METADATA
SUSPECT
FACE RECOGNITION MODULES
Library Classes Methods
• com.amazonaws.AmazonServi • face_rekognition • open_sketchActionPerform
ceException; ed(evt)
• com.amazonaws.*; • upload_sketchActionPerfor
• com.amazonaws.services.reko med(evt)
gnition • find_matchActionPerforme
• java.awt.*; d(evt)
• AmazonRekognitionClient
• com.fasterxml.jackson;
Builder.defaultClient()
• AmazonS3ClientBuilder.sta
ndard().withRegion(clientR
egion).build()
Advancements
TECHNOLOGIES
IMPLEMENTED
01 AMAZON S3
04 JACKSON(JSON PROCESSING)
02 AMAZON REKOGNITION
05 JAVAFX
03 JAVA SWING
THANK YOU!
