Facial Recognition and Verification with Mobile Devices.

Facial Recognition and Verification with Mobile Devices

Table of Contents

• Project Overview• Who Cares?• Current Progress• Future Plans• Incorporated Technologies

Table of Contents

• Project Overview• Who Cares?• Current Progress• Future Plans• Incorporated Technologies

Project Overview

Today, mobile devices are equipped with cameras that are more than 4x better than just 7 years ago.

iOS device, in particular, also have thousands of compatible accessories ranging from music players to 3D sensors.

iPhone 5S:8 megapixelsAutofocusHybrid IR filtersLED FlashIllumination SensorsFive-Element Lensf/2.4 aperture1080p Video Capture

Original iPhone:2 megapixels

Project Overview

Buildings and Homes all across the country use a wide array of complex security devices to protect against unwarranted access.

Why not use everyday devices? My project involves using an iPad and off-the-shelf depth sensor to perform 3D Face Recognition and Verification for security access.

Project Overview

My project is designed to replace the current key-card access system in the CBL lab (left) with my iPad/Sensor System (right).

Mock-Up

Table of Contents

• Project Overview• Who Cares?• Current Progress• Future Plans• Incorporated Technologies

Who Cares?

Why should everyone in the lab always remember to have their IDs on them?

• Get up to talk to someone who just walked by and the door closes…Locked out.

• Forget your wallet at home…Locked out.• Realize your meeting is soon and leave behind your

wallet in the rush…Locked out.

This project was inspired by the fact that there has to be a BETTER WAY.

Table of Contents

• Project Overview• Who Cares?• Current Progress• Future Plans• Incorporated Technologies

Current Progress

The most important thing for these 10 weeks is to have something to show for it.

Step 1: Create a System which uses Face Detection to unlock the lab door.

Step 2: Scan members of the lab to record face data.

Step 3: Incorporate a verification system which compares new face data to already recorded face data.

Current Progress

iOS Application (Face Detection)

Sample Scan(Structure Sensor)

Current Progress

Step 1: Create a System which uses Face Detection to unlock the lab door. - DONE

Step 2: Scan members of the lab to record face data. – IN PROGRESS

Step 3: Incorporate a verification system which compares new face data to already recorded face data.

Table of Contents

• Project Overview• Who Cares?• Current Progress• Future Plans• Incorporated Technologies

Future Plans

Complete the next step of the project outline.

There are many different conditions that can effect the performance and accuracy of facial scans.

Bad Conditions:• Different Pose• Dim Lighting• Background Noise• Etc.

Table of Contents

• Project Overview• Who Cares?• Current Progress• Future Plans• Incorporated Technologies

Incorporated Technologies

• iOS SDK• Structure Sensor SDK• OpenCV• OpenBR