Final Report Summary - INGRESS (Innovative Technology for Fingerprint Live Scanners)
Executive Summary:
The goal of INGRESS was to deal with the problem of damaged and altered fingers, preventing sensors to capture a fingerprint image of sufficient quality to perform fingerprint recognition. INGRESS developed 2 innovative technologies to capture internal fingerprints, i. e. fingerprint under the surface of the skin: ultrasound and Optical Coherence Tomography (OCT). Those sensors capture the fingerprint pattern present in the epidermis to get a fingerprint image that is usable for fingerprint recognition with legacy systems and algorithms. A third technology has been developed: a slim sensor based on Printed Organic Electronics (POE). This has several advantages compared with current sensors: low-cost and very small form factor even for large surface.
We started from scratch with the development of these new technologies. Thus, the challenges were high. For internal fingerprint, the OCT provided the best result. We demonstrated our capability to capture sub-surface fingerprint images with good quality that are usable for processing. We also showed that, in case of a damages fingerprint, the internal fingerprint image is able to recover the entire fingerprint compared to the external fingerprint. At the end of the project, while we made tremendous progress, we cannot pretend to outperform current solutions yet as there is still some work required to improve the quality and the accuracy of the technologies, but the OCT technology is quite close. Regarding POE, the objective for fingerprint recognition was to reach 500 dpi resolution. We did not succeed to finally build a working sensor at this resolution but there was just one step left to achieve a mock-up, so the project at least paved the way to the technology.
We also proposed new algorithms and methods for legacy sensors to deal with altered fingerprints with improved results compare to SOA.
Finally, an ethical study was conducted to provide Privacy-By-Design guidelines to be applied during the technology development.
Project Context and Objectives:
Project context
For over ten years, the use of fingerprints has been a key policy technology to help address both identification and security issues around the world. Unfortunately, the overall success of fingerprints for identification and verification purposes greatly depends on the quality of the fingerprints initially enrolled and acquired.
In Europe, this development was kicked off with the EURODAC Regulation (2000) to develop a EU centralised automated fingerprint identification system storing fingerprints of all asylum seekers. EURODAC went live in 2003. The European Visa Information System (EU VIS) was established in 2004 by Council decision 2004/512/EC and went live on October 11th 2011. The EU VIS is a database containing information, including biometrics, on visa applications by Third Country Nationals requiring a visa to enter Schengen Member States (MS). New applicants for a Schengen visa have to travel to the nearest EU consulate to give their biometric information (10 fingerprints and a facial image), which is then stored to support a fingerprint verification of the visa holder when arriving at a Schengen Border. In the US, after 9/11 US VISIT initiated the capturing of biometrics at US Borders in 2004.
Significant efforts have also been invested to make travel documents more secure by issuing passports with biometric data and embedded electronic chips (the “ePassport”). On both sides of the Atlantic, aggressive timelines have pushed for the adoption of ePassports: the PATRIOT Act in USA, and Regulation 2252/2004 in the EU. This regulation requested the Schengen Member States to deliver ePassports with facial biometric data from August 2006 onwards. Later, to strengthen security and trust, the EU introduced fingerprint biometrics as a mandatory feature for Schengen Member States which 27 Member States now meet. In the meantime about 310 million ePassports have been issued in 89 countries around the world and over 130 million in the EU alone.
Since 2006, the Directive 2006/126/EC recasts the existing legislation harmonising the conditions for issuing national driving licences, to make it easier for people to move within the EU and to reduce the scope for fraud. In 2013, France will apply this directive by the deliverance of new driving licence with biometric data. Certainly, this choice will spread in Europe in the future.
In view of the important role of biometrics to hinder criminal and terrorist movements, address immigration issues and hence contribute to the protection of civil society (identity theft, fraud...) there is an urgent need for a multi-disciplinary research and development initiative, validating innovative technologies on fingerprint live scanners. Due to the growing use of Automated Fingerprint Identification Systems (AFIS), it is expected that voluntary alterations of fingerprints will increase in the years to come. For the same reason (AFIS development), it can be expected that involuntary alterations, which are uncommon in the main population but are not negligible, will have to be addressed in order to give an equal access to biometric systems for all citizens. These initiatives have to focus on the quality of fingerprint images especially for the people with superficial skin disorders and on decreasing sensors weight, size and costs.
Scientific and technical objectives
The project focuses on capturing sub-surface fingerprint and delivering a high-quality image. The technology stream of the project focuses on medical imaging technique, such as ultrasound and Full Field Optical Coherence Tomography (FFOCT), to acquire the fingerprint matrix in the dermis. Furthermore, INGRESS studies the use of Printed Organic Electronics (POE) technologies and components, such as the passive matrix of Organic Light Emitting Diodes (OLEDs) and Organic Photo Diodes (OPDs) to create a new generation of high resolution and mobile fingerprint sensors.
As a side effect of looking beyond the finger surface, the new scanners will be able to better detect voluntary finger alterations (e.g. skin switching or transplantation), which are a critical issue in border control scenario, where several cases of illegal immigrants altering their fingers have been reported. Moreover, these technologies will overcome involuntary alterations and allow giving an equal access to the biometric technology for all citizens.
INGRESS’ mock-ups will be evaluated in laboratory to compare them with legacy sensors. Tests in the field will assess the technology performance, the usability and user acceptance of the solution. INGRESS will use both standard approaches and novel metrics to evaluate the operational quality of fingerprint scanners.
In parallel of the technology stream, the project will investigate the potential ethical, legal and societal issues for these novel technologies. The R&D process will integrate, all along the project, the results from this investigation, to guarantee that all end-products will be fully compatible/compliant with current trends in European and international privacy and data protection standards, following the internationally-endorsed foundational principles of Privacy by Design.
INGRESS will both validate the developed technologies and propose a technology development roadmap for the purpose of using fingerprints from identity documents in border control and law enforcement applications.
Project Results:
The goal of INGRESS was to deal with the problem of damaged and altered fingers. In case of altered fingerprints, the existing fingerprint sensors are not able to capture a fingerprint image of sufficient quality for automatic processing in order to perform fingerprint recognition. Therefore INGRESS projects developed 2 innovative technologies to capture internal fingerprints, i. e. fingerprint under the surface of the skin: ultrasound and Optical Coherence Tomography (OCT). Indeed, some studies showed that the fingerprint pattern in present in the epidermis. The goal of those innovative sensor is to capture this pattern and get a fingerprint image that is usable for fingerprint recognition with legacy systems and algorithms. Besides, a third technology has been developed in INGRESS: a slim sensor based on Printed Organic Electronics (POE). This has several advantages compared with current sensors: low-cost and very small form factor even for large surface.
In INGRESS, we started from scratch with the development of these new technologies. Thus the challenges were high. For internal fingerprint, the OCT technology is the one that provided the best result. We demonstrated with this technology our capability to capture sub-surface fingerprint images with good quality that are usable for processing. We also showed that, in case of a damages fingerprint, the internal fingerprint image is able to recover the entire fingerprint compared to the external fingerprint. At the end of the project, while we made tremendous progress, we cannot pretend to outperform current solutions yet as there is still some work required to improve the quality and the accuracy of the technologies. In the benchmark we did, the legacy sensors are still providing better accuracy results but the OCT technology is quite close. Regarding POE, the objective for fingerprint recognition was to reach 500 dpi resolution for the sensor while the state of the art was less 64 dpi. We did not succeed within the project to finally build a working sensor at this resolution but there was just one step left to achieve a mock-up, so the project at least paved the way to the technology.
The project also focussed on the development of algorithms to improve the matching of altered fingerprints captured with legacy sensors. We successfully proposed several algorithms and methods in order to minimize the influence of the altered sections of fingerprints images and to increase the accuracy of recognition.
Potential Impact:
Market applications
The initial use case for the technology was border control. Border police are facing a number of situation where it is very difficult to capture fingerprints of the traveller because of damages fingerprint (either voluntary or involuntary). So, border control would be a typical market application for sub-surface sensors. POE technology, would be capable when mature, to have large surface and thin sensors. This is touchscreen technology where you can basically transform the screen of a smartphone or in tablet in a sensor. Therefore, the mobile market would be the perfect target, but also any application requiring fingerprint sensors (access control, etc.)
Ethical, legal and societal issues
During this project, our ethical partner, conducted a study in order to propose and provide Privacy-By-Design guidelines to be applied during the technology development. This methodology is described in a document whose purpose is to help developers to take into account privacy when developing a new sensor.
Industry perspectives
The technology is not completely matured yet. Thus, we have no feedback so far from industry. From our perspective as IDEMIA (former Morpho) and coordinator of the project, OCT and POE technologies are very promising but need more development to reach an industrial phase. The technology however raised a lot of attention during various presentation in conference and public events. The first commercial applications are not for today as we have first to continue the work on the technology to bring it to a higher level of maturity. Therefore, plans will focus on this first.
List of Websites:
www.ingress-project.eu
The goal of INGRESS was to deal with the problem of damaged and altered fingers, preventing sensors to capture a fingerprint image of sufficient quality to perform fingerprint recognition. INGRESS developed 2 innovative technologies to capture internal fingerprints, i. e. fingerprint under the surface of the skin: ultrasound and Optical Coherence Tomography (OCT). Those sensors capture the fingerprint pattern present in the epidermis to get a fingerprint image that is usable for fingerprint recognition with legacy systems and algorithms. A third technology has been developed: a slim sensor based on Printed Organic Electronics (POE). This has several advantages compared with current sensors: low-cost and very small form factor even for large surface.
We started from scratch with the development of these new technologies. Thus, the challenges were high. For internal fingerprint, the OCT provided the best result. We demonstrated our capability to capture sub-surface fingerprint images with good quality that are usable for processing. We also showed that, in case of a damages fingerprint, the internal fingerprint image is able to recover the entire fingerprint compared to the external fingerprint. At the end of the project, while we made tremendous progress, we cannot pretend to outperform current solutions yet as there is still some work required to improve the quality and the accuracy of the technologies, but the OCT technology is quite close. Regarding POE, the objective for fingerprint recognition was to reach 500 dpi resolution. We did not succeed to finally build a working sensor at this resolution but there was just one step left to achieve a mock-up, so the project at least paved the way to the technology.
We also proposed new algorithms and methods for legacy sensors to deal with altered fingerprints with improved results compare to SOA.
Finally, an ethical study was conducted to provide Privacy-By-Design guidelines to be applied during the technology development.
Project Context and Objectives:
Project context
For over ten years, the use of fingerprints has been a key policy technology to help address both identification and security issues around the world. Unfortunately, the overall success of fingerprints for identification and verification purposes greatly depends on the quality of the fingerprints initially enrolled and acquired.
In Europe, this development was kicked off with the EURODAC Regulation (2000) to develop a EU centralised automated fingerprint identification system storing fingerprints of all asylum seekers. EURODAC went live in 2003. The European Visa Information System (EU VIS) was established in 2004 by Council decision 2004/512/EC and went live on October 11th 2011. The EU VIS is a database containing information, including biometrics, on visa applications by Third Country Nationals requiring a visa to enter Schengen Member States (MS). New applicants for a Schengen visa have to travel to the nearest EU consulate to give their biometric information (10 fingerprints and a facial image), which is then stored to support a fingerprint verification of the visa holder when arriving at a Schengen Border. In the US, after 9/11 US VISIT initiated the capturing of biometrics at US Borders in 2004.
Significant efforts have also been invested to make travel documents more secure by issuing passports with biometric data and embedded electronic chips (the “ePassport”). On both sides of the Atlantic, aggressive timelines have pushed for the adoption of ePassports: the PATRIOT Act in USA, and Regulation 2252/2004 in the EU. This regulation requested the Schengen Member States to deliver ePassports with facial biometric data from August 2006 onwards. Later, to strengthen security and trust, the EU introduced fingerprint biometrics as a mandatory feature for Schengen Member States which 27 Member States now meet. In the meantime about 310 million ePassports have been issued in 89 countries around the world and over 130 million in the EU alone.
Since 2006, the Directive 2006/126/EC recasts the existing legislation harmonising the conditions for issuing national driving licences, to make it easier for people to move within the EU and to reduce the scope for fraud. In 2013, France will apply this directive by the deliverance of new driving licence with biometric data. Certainly, this choice will spread in Europe in the future.
In view of the important role of biometrics to hinder criminal and terrorist movements, address immigration issues and hence contribute to the protection of civil society (identity theft, fraud...) there is an urgent need for a multi-disciplinary research and development initiative, validating innovative technologies on fingerprint live scanners. Due to the growing use of Automated Fingerprint Identification Systems (AFIS), it is expected that voluntary alterations of fingerprints will increase in the years to come. For the same reason (AFIS development), it can be expected that involuntary alterations, which are uncommon in the main population but are not negligible, will have to be addressed in order to give an equal access to biometric systems for all citizens. These initiatives have to focus on the quality of fingerprint images especially for the people with superficial skin disorders and on decreasing sensors weight, size and costs.
Scientific and technical objectives
The project focuses on capturing sub-surface fingerprint and delivering a high-quality image. The technology stream of the project focuses on medical imaging technique, such as ultrasound and Full Field Optical Coherence Tomography (FFOCT), to acquire the fingerprint matrix in the dermis. Furthermore, INGRESS studies the use of Printed Organic Electronics (POE) technologies and components, such as the passive matrix of Organic Light Emitting Diodes (OLEDs) and Organic Photo Diodes (OPDs) to create a new generation of high resolution and mobile fingerprint sensors.
As a side effect of looking beyond the finger surface, the new scanners will be able to better detect voluntary finger alterations (e.g. skin switching or transplantation), which are a critical issue in border control scenario, where several cases of illegal immigrants altering their fingers have been reported. Moreover, these technologies will overcome involuntary alterations and allow giving an equal access to the biometric technology for all citizens.
INGRESS’ mock-ups will be evaluated in laboratory to compare them with legacy sensors. Tests in the field will assess the technology performance, the usability and user acceptance of the solution. INGRESS will use both standard approaches and novel metrics to evaluate the operational quality of fingerprint scanners.
In parallel of the technology stream, the project will investigate the potential ethical, legal and societal issues for these novel technologies. The R&D process will integrate, all along the project, the results from this investigation, to guarantee that all end-products will be fully compatible/compliant with current trends in European and international privacy and data protection standards, following the internationally-endorsed foundational principles of Privacy by Design.
INGRESS will both validate the developed technologies and propose a technology development roadmap for the purpose of using fingerprints from identity documents in border control and law enforcement applications.
Project Results:
The goal of INGRESS was to deal with the problem of damaged and altered fingers. In case of altered fingerprints, the existing fingerprint sensors are not able to capture a fingerprint image of sufficient quality for automatic processing in order to perform fingerprint recognition. Therefore INGRESS projects developed 2 innovative technologies to capture internal fingerprints, i. e. fingerprint under the surface of the skin: ultrasound and Optical Coherence Tomography (OCT). Indeed, some studies showed that the fingerprint pattern in present in the epidermis. The goal of those innovative sensor is to capture this pattern and get a fingerprint image that is usable for fingerprint recognition with legacy systems and algorithms. Besides, a third technology has been developed in INGRESS: a slim sensor based on Printed Organic Electronics (POE). This has several advantages compared with current sensors: low-cost and very small form factor even for large surface.
In INGRESS, we started from scratch with the development of these new technologies. Thus the challenges were high. For internal fingerprint, the OCT technology is the one that provided the best result. We demonstrated with this technology our capability to capture sub-surface fingerprint images with good quality that are usable for processing. We also showed that, in case of a damages fingerprint, the internal fingerprint image is able to recover the entire fingerprint compared to the external fingerprint. At the end of the project, while we made tremendous progress, we cannot pretend to outperform current solutions yet as there is still some work required to improve the quality and the accuracy of the technologies. In the benchmark we did, the legacy sensors are still providing better accuracy results but the OCT technology is quite close. Regarding POE, the objective for fingerprint recognition was to reach 500 dpi resolution for the sensor while the state of the art was less 64 dpi. We did not succeed within the project to finally build a working sensor at this resolution but there was just one step left to achieve a mock-up, so the project at least paved the way to the technology.
The project also focussed on the development of algorithms to improve the matching of altered fingerprints captured with legacy sensors. We successfully proposed several algorithms and methods in order to minimize the influence of the altered sections of fingerprints images and to increase the accuracy of recognition.
Potential Impact:
Market applications
The initial use case for the technology was border control. Border police are facing a number of situation where it is very difficult to capture fingerprints of the traveller because of damages fingerprint (either voluntary or involuntary). So, border control would be a typical market application for sub-surface sensors. POE technology, would be capable when mature, to have large surface and thin sensors. This is touchscreen technology where you can basically transform the screen of a smartphone or in tablet in a sensor. Therefore, the mobile market would be the perfect target, but also any application requiring fingerprint sensors (access control, etc.)
Ethical, legal and societal issues
During this project, our ethical partner, conducted a study in order to propose and provide Privacy-By-Design guidelines to be applied during the technology development. This methodology is described in a document whose purpose is to help developers to take into account privacy when developing a new sensor.
Industry perspectives
The technology is not completely matured yet. Thus, we have no feedback so far from industry. From our perspective as IDEMIA (former Morpho) and coordinator of the project, OCT and POE technologies are very promising but need more development to reach an industrial phase. The technology however raised a lot of attention during various presentation in conference and public events. The first commercial applications are not for today as we have first to continue the work on the technology to bring it to a higher level of maturity. Therefore, plans will focus on this first.
List of Websites:
www.ingress-project.eu