Despite the advancements in fingerprinting recognition technology and availability of highly accurate fingerprint readers, a dirty finger or smudge on the scanning surface is enough to disrupt the performance of your scanner. In some less common cases, technical limitations and inefficiencies may be the causes of poor fingerprint capture. However, in most cases, improper use or poor maintenance is the reason of poor scanner performance.
In this article, we will discuss about different fingerprinting techniques, how to properly use a fingerprint scanner and keep its performance up to the mark. We will also discuss how can you get clear fingerprint image by keeping a few things in mind and will share some fingerprint scanner maintenance tips.
Fingerprinting has a long history and experts have developed several fingerprinting techniques for capturing, processing and matching of human fingerprints. Back when there were no IAFIS systems, and not even computers or scanners, there were manual fingerprinting techniques for the identification of individuals. However, fingerprinting was mostly used in forensics for the identification of criminals or for solving the crime mysteries.
Ink based fingerprinting
This is the oldest, easiest, cheapest yet a reliable fingerprinting technique, which has been used since the medieval period. On ancient Babylonian clay tablets, fingerprint impressions were found, which were used as identification marks by the potters. In forensics and law enforcement applications, fingerprints impressions are taken by applying a layer of ink on the fingertip and rolling against the fingerprint card. Fingerprint impressions have been also used for legal documents like agreements, deeds, etc.
Ink based fingerprint technique is still relevant in many applications including law enforcement, in which fingerprints impressions are captured using ink during criminal booking. However, in most digital applications, ink based fingerprinting technique offers no help. Law enforcement agencies like FBI accept ink based fingerprint impressions for biometric background check. However, these paper based fingerprints are scanned to create a digital image and run a match. In most digital applications the more advanced and efficient fingerprinting technique, digital fingerprinting is used.
Electronic / digital fingerprinting techniques
Ink based acquisition and manual processing of fingerprints has served human identification needs very well and is still a tried-and-true way to capture fingerprints. However, ink based fingerprinting is slow and does not fit in today’s fast paced digital needs. Time taken in capturing and processing fingerprints is horrendously high,which only suits specific applications. Need of faster fingerprint data acquisition and processing paved the way to electronic / digital fingerprinting, which has already become the most popular biometric modality and a mainstream method for personal identification.
Several techniques have been developed for digital fingerprinting. They can be put in different categories on the basis of factors like sensing methods, captured dimensions, etc.
Electronic / digital fingerprinting techniques by sensing methods
In any electronic fingerprinting device, sensor is an important piece of hardware, which is responsible for capturing the fingerprint image. There may be different sensing methods depending on physical and chemical properties of the material used.
Oldest, yet reliable fingerprint sensing method. Early fingerprint recognition systems made use of optical imaging to capture fingerprint details and it is still in use. It is the simplest setup that uses visible light to capture fingerprint details with the help of a digital camera and a prism that enables total internal reflection. Optical sensors are cheap and are used in a low to high-end fingerprint sensors to keep the cost down.
Unlike optical sensors, capacitive sensing makes use of pixel array of capacitors to acquire image of user fingerprints. Being small in size, they are very popular on mobile devices. Capacitive sensors are costlier than the optical sensors but cheaper than the ultrasonic and thermal sensors.
This sensing method uses very high frequency sound waves to capture the pattern of friction ridges on fingertips.
Thermal sensing makes use of the fact that there is a very slight temperature difference between ridges and valleys of fingerprints, which can be measured using thermal line sensors and generating the pattern image
Electronic / digital fingerprinting techniques by dimensions captured
2D fingerprinting techniques
Most scanners and sensors capture two dimensional image of fingerprint image with the help of different sensing techniques.
3D fingerprinting techniques
To overcome the problem of distortion and noise in fingerprints while touching and pressing against the scanning surface, three dimensional fingerprinting was introduced. 3D fingerprinting technique requires the subjects to roll the finger on the sensor surface. There are also fingerprinting systems that enable contact less capture of 3D fingerprints.
Electronic / digital fingerprinting techniques by level of intrusiveness
These fingerprinting techniques require the subject to touch the scanning surface.Most fingerprinting applications make use of contact based fingerprinting technique.
Touch less / contact less fingerprinting
Most fingerprint recognition systems require subjects to touch a scanning surface, which is considered intrusive. To reduce the level of intrusiveness, contact less fingerprinting techniques were developed in which fingerprint details can be scanned from a distance using a camera setup.
Electronic / digital fingerprinting techniques by the amount of details captured
In mobile and portable devices smaller sensors and components are used to keep the device lightweight and compact. These sensors and the underlying recognition algorithm are designed only to capture and process a portion of fingerprints.
Fingerprint recognition systems, in which compactness is not prime requirement, larger area of fingerprints is captured and process. Since full fingerprint systems capture and process way more details than the partial ones, they are considered securer.
How accurate is a fingerprint scanner?
It has been long established that each fingerprint is unique and no two fingers (even of the same person or even twins) carry similar fingerprint features. Now when we know that fingerprints are unique, we only have to concentrate on the technology part that how to capture, process and match these unique characteristics, so that there are virtually no errors. While keeping error rate to zero is practically not possible due to several reasons, present day fingerprint recognition systems are highly accurate and offer high level of tolerance against several environmental factors, which could lead to unreliable results otherwise.
This study, is however more subjective as it looks at the number of times the fingerprinting attempts were made. A more detailed approach would consider many other factors like encrypted domain or where or when the template protection is implemented.
Fingerprint accuracy is unlike password accuracy
We are well acquainted with password based authentication systems. Password based authentication provide a clear match or no-match. They are either 100 percent accurate or 100 percent inaccurate. Now biometric systems have started to take over in many authentication applications,where earlier passwords or PINs were used. Going through this transition, people often tend to compare biometric systems with passwords and unrealistically expect them to match with the accuracy of password based authentication systems.
If you are an enrolled user and presenting the accurate biometric identifier, a biometric system may not let you in, while a password based system always will. Biometrics is unlike passwords.
Scanning the correct finger on a biometric authentication system is not comparable with entering the correct password on a password based authentication system. Correct passwords will always be treated correct on password based systems; however, that is not the case with biometric systems. Even your accurate biometric scan can return a no-match and deny access.
Understanding fingerprint recognition accuracy
Biometric matching cannot be 100% accurate. The accuracy of a biometric system can be evaluated by using well-known performance indicators, e.g., False Accept Rate (FAR), False Reject Rate (FRR), and Equal Error Rate (EER). Recognition accuracy generally depends on factors such as input image quality and matching algorithms. Due to constant advancement in technology and research, biometric systems have been able to achieve remarkable matching accuracy.
There are evaluation platforms for biometric recognition algorithms, where researchers can submit biometric recognition algorithm developed by them and compete with other algorithm on matching accuracy. FVC-ongoing provides one of such web-based automated evaluation system for fingerprint recognition algorithms. Tests are carried out on a set of sequestered datasets and results are reported on-line by using well known performance indicators and metrics.
Individual researchers, companies as well as academic research groups have been able to report some ground breakingaccuracy stats with the recognition algorithms submitted by them. The best matching accuracy (as of April 2019) out of the fingerprint verification competition was EER = 0.022%, which was reported by Beijing Hisign Bio-info Institute and achieved with their HXKJ(version 2.4) algorithm. Algorithm named MM-FV version 11.0 was able to achieve Equal Error Rate of 0.027%. MM-FV was developed by a technology firm named Neurotechnology and was reported on FVC-ongoing in January 2019.The state-of-the-art MCC (Minutia Cylinder Code) based fingerprint matching algorithm achieved the EER = 0.49% on database FVC2002 DB2, and the EER = 0.12% on database FVC2006 DB2.
|27-7-17||FV-STD-1.0||Beijing Hisign Bio-info Institute||Company||HXKJ||2.4||0,022%||0,007%||0,036%|
|29-8-11||FV-STD-1.0||Tiger IT Bangladesh||Company||TigerAFIS||1.2ec||0,108%||0,115%||0,242%|
|14-9-10||FV-STD-1.0||Green Bit S.p.A||Company||GBFRSW||188.8.131.52||0,118%||0,155%||0,519%|
|31-8-11||FV-STD-1.0||AA Technology Ltd.||Company||EMB9300||1.1||0,142%||0,159%||0,220%|
Here, it is worth noting that above results are based on FVC-ongoing benchmark FV-STD-1.0, i.e. they were achieved in ideal conditions. FV-STD-1.0 contains fingerprint images acquired in operational conditions using high-quality optical scanners. Algorithm results using FV-STD-1.0 would reflect the expected accuracy in large-scale fingerprint-based applications.These algorithms were fed with better quality of fingerprint images than non-ideal conditions (for example, latent prints from crime scenes). If we have a look at the result using the same algorithm and benchmark FV-HARD-1.0, which contains a relevant number of difficult cases (noisy images, distorted impressions, etc.) and makes fingerprint verification more challenging.
|28-7-17||FV-HARD-1.0||Beijing Hisign Bio-info Institute||Company||HXKJ||2.4||0,530%||0,797%||1879%|
The decline in performance stats for top two algorithms is apparent. The performance drop proves that image quality is the most essential criteria, which can improve or distort the performance of a recognition algorithm.
Why do fingerprints get rejected?
Despite flaunting accuracy of more than 99 percent of the time, people do come across failed fingerprinting attempt in practical scenarios. There can be many instances a fingerprint recognition system will reject your correct fingerprinting attempts, when it finds that sample is not good enough to be processed and ask you to re-scan your finger.
This happens because external factors affect the quality of fingerprints more than the internal inefficiencies of the recognition systems. Avoiding the causes of poor fingerprints can help improve the fingerprint quality and help get a clear fingerprint image.
User behavior is one of the main causes of poor fingerprints. Improper user behavior during the fingerprint acquisition process is the prime reason why fingerprints get rejected. Improperly placed fingertip, dirty / too wet / too dry fingertip, unsteady finger placement may cause fingerprints to get rejected.
Poor scanner maintenance
Poor scanner maintenance and harsh usage can be called the #2 reason why your fingerprints attempts can fail. Today’s fingerprint scanners are adequately robust; however, they still require some maintenance and care to keep up. Cleaning scanner glass, keeping it away from dust and liquid, gentle usage and avoiding rubbing the scanner glass, etc. are some of the easy ways you can have the scanner serving you for years.
Poor quality fingerprint scanners
Price of fingerprint scanners very from $25 to $4500 and may even go lower or higher than that. A low end fingerprint scanner may be able to serve your purpose but you will have to make compromises in terms of accuracy, security and efficiency. Using low-end inexpensive parts, they may not produce high quality fingerprint images or capture enough details, making it more challenging for recognition algorithm to be efficient.
Several environmental factors can affect the performance of the recognition system and cause poor fingerprint images. A fingerprint scanner can reject your fingerprints in intolerable environmental conditions. Very high or low temperature, high level of humidity, extremely dusty, dry and wet environment, can cause poor fingerprints resulting in failed attempts.
Limitations of the recognition system
Today’s fingerprint scanners are adequately accurate; however they are still not 100% precise and have their own set of limitations. Other than that, quality of sensors very greatly from low to high-end scanners. Acquisition area and the output resolution are the most critical quality parameters, which, at the PIV IQS minimum requirements, cause an average performance drop of 73% and 20%, respectively.Limitations of hardware or software like Inefficiencies of data acquisition subsystem, matching algorithms, etc. can also be the causes of poor fingerprints.
Automated fingerprint recognition systems require a certain number of fingerprint features to establish and verify identity. A small percentage of target population may have unusable fingerprints for the purpose of automated identification. There may be several reasons behind this, e.g. certain skin diseases, injury, laborious work conditions, friction, etc. Unusable fingerprint is one of the causes of poor fingerprints and why fingerprints get rejected for some individuals.
How to get a clear fingerprint image and improve image quality?
We discussed have above that user behavior, low quality scanners and poor maintenance are the top causes of poor fingerprints, which leads to failed scanning attempts. There are a few things you can take care of to improve image quality and make sure that you always get a clear fingerprint image every time you scan your fingerprint.
Go for a good quality scanner
Scanner quality can make all the difference in matching accuracy, efficiency and security of a fingerprint recognition system. Investing on a good quality scanner is always a good idea specially if there are high expectations for security, accuracy and efficiency. Good quality scanners may be more expensive than low-end offerings, but they use good quality sensors, electronic components and material that results in improved performance, speed and accuracy.
Use the fingerprint scanner properly
Just getting a good quality scanner is not enough until you know how to properly use it. Fingerprint scanner can be a delicate device, specially the scanning area, which is mostly made of glass.
- Try to keep your fingertip clean whenever you get it scanned. Clean finger during the enrollment process will save you from a lot of frustration later.
- Know your scanning time: it may be instant or take a few seconds, depending on the amount of data it has to process. You will come to know it after a few attempts.
- Keep your fingertip steady (if it is a static scanner) or swipe it with a normal pace on swipe fingerprint readers.
- Read / listen to the error messages for any failed attempts, if you come across failed attempts frequently, you may have to go through the enrollment process again.
Once you are done with Invest-on-a-good-quality-scanner part, regular maintenance is all you need to do get the best results out of it.
Glasses are smudge magnets and once it gets a layer of smudge on it, it start affecting the performance of your scanner. Getting rid of dirt, dust and smudge is easy if you do it with right cleaning material.
How to clean scanner glass?
Step by step method for cleaning scanner glass
- Unplug your fingerprint scanner from the power source or from any device if it is connected.
- If there are visible or invisible fine dust particles, clean the scanner glass with a soft piece of cloth.
- If large particles of dust or dirt are there, do not try rubbing with cloth, instead use a light blower to blow them away.
- Damp a lint-free soft piece of cloth (preferably cotton) with glass-cleaner solution or acetone.
- Clean the scanner glass by gently rubbing the cloth over the glass surface.
- Repeat the process if required. Make sure there no residual solution on the scanner glass.
- If scanner glass still feels cloudy, damp a soft cloth with ammonia based cleaner and gently clean the glass until dried.
- To remove stubborn residues or cloudiness, place scotch tape on the scanner glass and peel it off.
- You can repeat scotch-tape step for better results.
- Done! Get your scanner back to work.
Things to take care of while cleaning the scanner glass
Do not use harsh chemicals or alcohol based cleaners to clean the glass surface.
Do not rub the glass surface abrasive material.
Do not rub the glass surface harshly as it may permanently scratch your glass surface.
Do not pour the glass cleaner directly on the scanner glass.
Scanner maintenance does not require any expert skills but the improvements in scanner performance after simple cleaning will surprise you.
Using a fingerprint scanner properly is as important as choosing the right fingerprint scanner. Investing on a high quality scanner is always a good idea as it will not only perform better but will give better return on investment in long run.Accuracy of a fingerprint scanner does not depend entirely on the hardware or the underlying algorithm. How you use it and maintain it, are equally important. So next time before you blame your fingerprint scanner for being inefficient or not performing adequately, ask yourself: when was the last time you cleaned it?