Touch-less Fingerprint: Can It Fill the Gaps Left by Standard Fingerprint Recognition?

While referring fingerprint recognition or fingerprint biometric systems, they are taken to be of “touch-based” by default. This is hardly surprising as most fingerprint recognition systems are of touch-type.They are the fingerprint recognition systems, in which subjects need to touch the sensor for biometric recognition to take place.

So, are there touch-less fingerprint recognition systems as well? Yes, there are. Arguably, most people may not even realize that touch-less method of fingerprint recognition exists. In terms of market share, touch-less fingerprint recognition is indeed non-existent; however, that was once the case with all the technologies popular today.

Increasing penetration of biometrics in identification and authentication applications opens a window of opportunity for touch-less fingerprint recognition. Touch-less fingerprint recognition is also expected to fill the gaps present in traditional touch-based fingerprint biometrics. Recent rise of smart phones has also paved the way for smart phone based touch-less fingerprint recognition due to several inherent advantages it offers.

Touch-less fingerprint recognition is an approach of biometric fingerprint recognition, in which subjects are recognized with their fingerprints but without making any physical contact with the recognition system.

Touch-less fingerprint recognition allows people to get their fingerprints scanned from a short distance (of a few inches) and they do not need to touch the sensor. In traditional touch-based fingerprint recognition, slight pressure is also applied while making the physical contact with the sensor. Pressure leads to undesired consequences such as elastic deformation. Since no physical contact is made, friction ridges of the scanned finger do not alter in shape as is the case with touch-based recognition systems.

There are several other advantages that touch-less fingerprint recognition offers. They are discussed in subsequent sections.

Smart phones have been working like a driving force behind the growth, adoption and social acceptance of biometric recognition. Before the inception of mobile devices with biometric capabilities, biometric recognition was largely limited for government and business applications.

Manufactures are now bringing cutting edge biometric recognition technologies to the smart phones. In this ever changing smart phone ecosystem, touch-less fingerprint recognition can be a refreshing change for the users.

Touch-less fingerprint recognition aims to leverage a sensing method that can capture fingerprint from a distance, unlike touch-based data acquisition technique found in traditional recognition systems. This touch-less data acquisition method should also be able capture fingerprints with high level of precision and detail, which is a primary requirement in any biometric recognition system.

Camera systems found on today’s smart phones check all the boxes and are widely used to accomplish touch-less fingerprint image capture.

The main difference between touch-based vs. touch-less fingerprint recognition is the method of capturing fingerprint details.

Using digital images to acquire fingerprint data is not unprecedented in fingerprint biometrics. Touch-based optical fingerprints recognition systems also capture fingerprint images digitally with CCD or CMOS image sensors. However, the sensor stays within the device. These devices leverage total internal reflection using a prism, in which a light beam reflects from the prism surface collecting the fingerprint image.

A digital camera can efficiently capture friction ridges from a distance. Fingerprint recognition on mobile devices is now extensively popular and these devices also come equipped with one or more cameras.The inbuilt digital cameras on these devices can play the role of data capture subsystem to perform touch-less fingerprint recognition. As camera systems on present day computing and mobile devices improve in terms of quality and resolution, performance of touch-less fingerprints recognition also improves accordingly.

A smart phone app can be devised to securely capture fingerprint image. A finger-shaped overlay on the digital viewfinder can be designed to guide the user about finger placement while collecting the print. Since the fingerprint image is securely captured by the app, it cannot be accessed by the user or any other app to ensure integrity.

Captured fingerprint image is taken though the pre-processing stage to enhance the image quality and make it useable for further processing. Pre-processing removes fingerprint image quality problems like non-uniform lighting by image normalization. The process of normalization calculates the mean and variance of a capture image and thus eliminates the difference in the illumination.  The image is then converted to a gray-scale image as smart phone camera systems capture images in RGB format.

Removal of unnecessary noise and background in the image along with reduction in size of the input image is also performed in the stage called Segmentation.

Post pre-processing, the enhanced fingerprint image is taken through fingerprint enhancement algorithm. Unlike pre-processing stage, which aims to enhance the quality of the captured image, fingerprint enhancement algorithm improves the clarity of ridge and furrow structures of input fingerprint image, based on estimated local ridge orientation and frequency.

Fingerprint matching can be done by matching the ridge valley structure. Mainly two types of techniques are used for fingerprint matching.

1. Minutiae based matching
2. Correlation based matching

Minutiae based systems try to identify the location and type of minutia and match it with the reference template. Accuracy is minutiae based systems is entirely dependent on the identification of minutia point. In case of correlation based techniques, the system performs global matching of ridge valley structure and tries to match the texture of fingerprint.

A touch-less fingerprint recognition system aims to capture and pre-process fingerprints locally, i.e. on device. However, depending on where the processed fingerprint template is matched, the system can be of the following types: Device based or Server based.

A device based system will match the processed fingerprint with reference template on-device. In device-based deployments, reference template is securely stored on the device itself. This type of deployment of touch-less fingerprint recognition is suitable for applications like device security, app authentication, e-commerce, banking, etc. It is easy to deploy and by installing a smart phone app and capturing fingerprint with device camera, the functionality can be enabled.

Server based touch-less fingerprint authentication is more suitable for applications like law enforcement, government applications, KYC, etc., in which a user identify has to be verified against what is on the database. In server based deployment, enrolled fingerprint templates are stored on a server and identity verification request can be placed remotely (e.g. from a smart phone).

This is the primary reason that paved the way to touch-less fingerprint recognition. Naturally, fingers are curved and full-fingerprint cannot be captured just by placing the finger against the flat sensor surface.On the other hand, when finger is placed on a touch-based fingerprint scanner, friction ridges deform due to the pressure and natural elasticity of the human skin.

One solution that is used is to roll and scan finger by capturing multiple frames per second. These frames can be used to create a single full-fingerprint template. However, this method does not completely solve the problem of elastic deformation and fingerprint recognition systems that offer this ability are expensive.

A touch-less fingerprint recognition system can effortlessly capture fingerprint without causing elastic deformation.

Touch based systems may contribute to spread infections and contagious diseases. It can become a medium to spread disease-causing bacteria and viruses as they rest on the fingerprint sensor surface and spread with touches. According to a report on 2019-20 coronavirus outbreak, the deadly corona virus could survive on surfaces like door handles and bus poles for 9 days. It is not hard to image how a fingerprint scanner touched by several people could have contributed to spread the disease.

Cost is an important factor when leveraging biometric recognition as biometric hardware can be a costly affair. Growing use of fingerprint authentication in all sorts of applications has helped it make more affordable and fingerprint recognition hardware is now cheaper than ever. However, it is still costlier than many other identification methods.

For example, to secure a device with PIN / password, developers just need to add a few lines of code, which does not introduce additional cost. However, to introduce fingerprint authentication on the same device, manufacturers of the device has to purchase additional hardware.

Touch-less fingerprint recognition is cheap to implement, you do not require any additional fingerprint hardware and existing on-device camera can be used to capture fingerprints.

Latent prints have helped solve numerous crimes as people at a crime scene can be identified with latent prints. However, they actually become a problem in civil applications and can impact the performance of a touch-based fingerprint scanner negatively.

On the other hand, touch based fingerprint scanning surface can be prone to accumulation of dirt, oil and other contaminants,mainly due to its repetitive contacts with different fingers with different levels of cleanliness. Coming in contact with skin contaminants accumulate some portion of them on the scanning surface. A touch-less sensor does not suffer from this issue.

Due to repetitive physical contact, dirt, dust and user behavior (e.g. rubbing the surface), the scanning surface of the touch-based fingerprint devices gets scratches life of the device is reduced. This is not the case with touch-less fingerprint recognition as it never comes in contact with the users.

Despite being less intrusive than many other biometric recognition methods, touch-based fingerprint recognition does involve some level of intrusiveness.Users are required to physically touch the fingerprint sensor to get identified / authenticated. A touch-less fingerprint recognition system can help reduce the intrusiveness.

Today’ mobile devices offer a lot of scope for smart phone based touch-less fingerprint recognition. Present day smart phones come equipped with one or more biometric recognition methods; and most of them include a fingerprint sensor. However, inbuilt fingerprint sensors found on today’s mobile devices have their limitations and have to make many compromises to become usable.

To maintain overall device compactness, highly compact fingerprint sensors are used on mobile devices. These sensors do a partial fingerprint scan, which means that only a small portion of fingertip is captured and matched. On the other hand, in most government, business and high security applications, systems that can process and match a full fingerprint scan, are required.

Owing to this limitation, fingerprint sensors found on smart phones cannot be deployed in applications which require a full-fingerprint scan.

Though most smart phones come equipped with a dedicated fingerprint sensor, a large portion of them (low-end smart phones) is still manufactured without it to keep the prices down. However, even the low-end smart phones are launched with one or more cameras. These smart phones can make use of touch-less fingerprint recognition by enabling the ability through an app or within the OS.

Touch-less fingerprint recognition can help capture full fingerprint with device camera and can be used for applications like FBI, law enforcement, KYC, etc. It can be usable on smart phones with or without inbuilt fingerprint sensors. In smart phones without an inbuilt fingerprint sensor, touch-less fingerprint recognition can also take care of device security, authentication and more.

Fingerprint biometrics has become a mainstream method of identification in numerous applications; however, mainly touch-based biometrics is used for the purpose. Smart phone based touch-less fingerprint recognition is based on the idea that most smart phones includes a camera, which can be used to read user fingerprints with a secure mobile app.

Touch-less fingerprint biometrics is a promising technology and can patch many gaps present in touch-based fingerprint biometrics. However, it also carries its own set of shortcomings. Accuracy, standardization and security are some of the concerns that need to be worked upon. Liveness detection is a major challenge in touch-less fingerprint recognition. While touch-based fingerprint recognition systems can take advantage of physical contact with the human skin and can check many biological or physiological factors to confirm liveness, touch-less fingerprint recognition does not have this advantage.

There are many challenges that touch-less fingerprint biometrics have to face and conquer.

As of now, only a few products are commercially available that offer smart phone based touch-less fingerprint recognition technology. However, with adequate funding, research and right marketing, touch-less fingerprint biometrics can also gain the level of ubiquity that touch-based fingerprint recognition enjoys.