Secure Human Identification Using Fingertip Biometrics Based on Ultrasound Rainer M Schmitt, PhD, CTO Sonavation Inc.

Talk Charleston

Sonavation Proprietary

Introduction • The ever increasing role of mobile devices in everyone’s life and the dominant IoT

• Threat to everyone due unauthorized use, fraude activities even terror attacks.

• In particular mobile devices are considered to be very easy to attack • fingerprint sensors are in favor due to their convenient use combined with a very high individual signature to be used as password and pin.

Sonavation Proprietary

Most Common Passwords in 2014

About 2.5 % of all devices were protected by these 25 passwords

Source:http://www.networkworld.com/article/2872085/micro soft-subnet/25-most-commonly-used-and-worst-passwords-of -2014.html

Sonavation Proprietary

Introduction • Biometric authorization for mobile devices, computers and networks is considered highly effective

• Optical, capacitance (electric-field) and ultrasound sensors have been developed for fingerprinting • While optical devices are bulky and expensive, capacitance sensors can easily be spoofed as recently demonstrated. Sonavation Proprietary

Advantage of Ultrasound Ultrasound is highly sensitive to fingerprint structure of ridges and valleys

Method

Optical

Capacitive, E-Field

Ultrasound

Parameter Sensed

Refractive Index nr

Permittivity εr

Spec. Acoustic Impedance Zsp [Rayl]

Ridge Valley Ratio

1.4 : 1

20:1

3750:1

In addition ultrasound is able to penetrate tissue and image under skin morphology

Sonavation Proprietary

Unspoofable Identity Over the last decade we have developed and we step wise implement a concept, which combines ultrasonic fingerprinting with subcutaneous imaging to obtain a highly secure individual signature which is almost impossible to spoof.

UNSPOOFABLE IDENTITY Fingerprint

Subcutaneous Biometrics Sonavation Proprietary

Introduction •

The technological challenge hereby is to create a sensor which is capable of robust, high resolution (500 dpi) fingerprinting while being able to image subcutaneous tissue with sufficient contrast and resolution. –

This technical concept is being presented and current status of its implementation being demonstrated

–

Principle of acoustic impediography for fingerprinting

–

Results

–

Imaging fingerprints through layers (notably glass)

–

First results

–

Minutiae based fingerprint identity

–

Fingerprinting through layers

–

Turning the fingerprint sensor into a 3D imaging system for anatomy, imaging tissue characteristics and blood flow

–

Test beds and first results

Sonavation Proprietary

Acoustic Impediography Acoustic Load Impedance Z = √rY r = Density Y = Young’s Mod Electrode

PZT Pillar

Valley: free ringing bell

Polymer

Ridge: bell ringing damped

Sonavation Proprietary

AC

Voltage Source

Fingerprint Imaging Sensor Substrate 1-3 Piezo Composite

Matrix density defines resolution. Pillars per inch = Dots per inch PPI=DPI

PZT Pillars embedded in Polymer

Sonavation Proprietary

Basic Principle of Operation

Sonavation Proprietary

Addressing Scheme Since 1000’s of pillars are hard to be addressed individually a cross-hatched pattern of electrode lines is used for interconnecting pillars to peripheral electronics.

Sonavation Proprietary

How Acoustic Impediography is implemented

Valley Impedance at fp

Finger

Ridge

Impedance

Ridge Impedance at fp

Valley

Ridge Impedance at fs

Fingerprint Sensor

Valley Impedance at fs

fs Series Resonant frequency

fp

Valley Transmitted Wave Fingertip Ridge

Frequency

Ridge Transmitted Wave

Fingertip Valley

Parallel Resonant frequency

Valley Current at fs Current

Current

Ridge Current at fs

Mechanical Resonator Sensing Element

Time

Start of Transmit cycle

Time

Start of Transmit cycle

Start of Sampling Current Buildup period

Start of Sampling Current Buildup period

Drawing: C. Liautaud

Sonavation Proprietary

Sensor Topography C um n

n

n

n

n

n

um

um

um

ol

ol

ol

C

C

ol

um

um

ol

ol

C

C

C

M

5

4

3

2

1

Sensor Array of MxN elements

Transmitters Row 1 Row 2 Row 3 Row 4 Row 5 Mechanical Resonator Sensing Elements Row N-1 Row N

Transmit Lines Receive Lines

Mechanical Resonators Sensing Elements Receivers

Selected Mechanical Resonator Sensing Element

Transmitter

Current measured

Current flow

Receiver Input

Adjustable Current-to-Voltage Amplifier

Noise Filter

Signal Conditioning

Adjustable Gain & Offset

Analog-toDigital Converter

Sonavation Proprietary

Data Storage

Data Processing Unit

Fingerprint

Sonavation Sensors

Swipe Sensor

Touch Sensor

Sonavation Proprietary

Comparison Optic Versus Acoustic

Sonavation Proprietary

Biometric Image Processing

Raw Image

Image enhance ment

Histogram Equalization Ridge Gradient enhancement (e.g. Sobel Filter) Ridge orientation estimate Segmentation (Border remo ve) Ridge Wavelength estimate Filter (Gabor,Median,Gauss) Binarizing

Minutiae Extraction Template Generation

Ridge thinning Minutiae extraction Template Generation

Template Matching

You are you!

Fingerprinting Through Layers • Mobile device mfg’s are highly interested to mount FPsensors underneath glass.  We expanded the contact impediography for under glass fingerprinting. Under glass: reduced resolution and contrast Glass layer is an additional load reduced contrast Wave propagation through layer diffraction limited, i.e. lower resolution

Sonavation Proprietary

Fingerprinting Through Layers • Despite reduced image quality biometric content is almost preserved

Sonavation Proprietary

Fingerprint and Minutiae Extraction Optical sensor

Touch sensor No Glass

Sonavation Proprietary

Touch sensor 0.4-mm Glass

3D/4D Ultrasonic Biometrics

3D/4D scan of subcutaneous tissue

3D/4D ultrasound echo data

Sonavation Proprietary

Superficial Fingerprint Vs. Subcutaneous Features

Sonavation Proprietary

3D Volume Imaging

Sonavation Proprietary

3D Volume Imaging

Sonavation Proprietary

Blood Flow Imaging

Sonavation Proprietary

Blood Flow Imaging

Sonavation Proprietary

Blood Flow Mapping

Sonavation Proprietary

Conclusion: •

The fingerprint swipe sensor based on acoustic impediography has been successfully released to the market on April 24, 2015

•

The corresponding touch sensor is in its final phase to be released by this year’s end.

•

Including subcutaneous biometrics is the scheduled for next year. This device is a medical grade instrument allowing us to define biometric standards for combined surface fingerprint and individual fingertip anatomy.

•

We are also exploring the use of a slightly modified sensor for skin cancer diagnostics and blood pressure monitoring

Sonavation Proprietary