Fingerprint technology is one of the most popular biometric modality to verify the identity of individuals. Fingerprints were first used in China in 700 AD and no two fingerprints have been found to be identical so far. Fingerprints are unique to even identical twins which make fingerprint biometrics a highly accurate and reliable identity verification method. Fingerprint matching compares the unique features such as the characteristics of ridges or minutia patterns that are found within the print pattern.

The fingerprint sensing process typically consists of capturing the fingerprint image, extracting the distinguishing features of the fingerprint, and then storing a digital template of the fingerprint or comparing the current image with the stored fingerprint templates.

An electronic device that records a digital image of the fingerprint pattern is known as a fingerprint reader. The captured image is known as a live scan which is then digitally processed. The distinguishing features are extracted and a fingerprint biometric template is created. This biometric template is stored and will be used for matching later.

A fingerprint scanner has many advantages over traditional identification mechanisms and provides businesses with a higher level of security. Organizations experience various benefits by incorporating a fingerprint reader such as reliable background checking of employees, secured access to facilities and assets and protection of confidential data. There are many types of fingerprint sensor technologies. We will be discussing the following three types of sensors namely optical fingerprint, capacitive based fingerprint and multispectral imaging.

Optical scanners are the common types of fingerprint scanners that use an LED light to illuminate the finger. The sensor detects and creates the fingerprint image by determining the light and dark areas created by the fingerprint ridges. The scanning process starts when the individual places his finger on the glass plate that is known as a touch surface.

Optical sensors mostly use two types of detectors – charge-coupled-devices (CCD) and CMOS based optical imagers. The CCD detectors are very sensitive to low light levels and are thus able to make excellent grayscale pictures. An inverted image of the finger is generated where the darker areas represent more reflected light and the lighter areas represent less reflected light. The darker areas are actually ridges of the finger and the lighter areas are the valleys between the ridges.

Prior to comparing the individual’s fingerprint with the stored template, the scanner processor ensures that a clear image has been captured by the CCD. It checks for various attributes such as the average pixel darkness or the overall values in a sample. The scan will be rejected by the processor if the overall image is too dark or too light. In such a case, the exposure time is adjusted by the scanner to allow more or less light and will try to scan the fingerprint again. When the darkness level is adequate, the scanner will check the image definition to determine the sharpness of the fingerprint scan. The scanner processor looks for many horizontal and vertical straight lines moving across the image. A line that runs perpendicular to the ridges and made up of sections that alternate between very dark and very light pixels implies that the fingerprint image has a good definition. The scanner processor proceeds to compare the captured fingerprint with the stored template if it finds that the captured image is crisp and properly exposed.

Charge coupled devices are however relatively expensive to fabricate and moreover fingerprint recognition does not require low-light sensitivity or grayscale imaging. On the other hand, CMOS based optical imagers can be manufactured at a lower cost. The CMOS imagers are manufactured in quantity and some of the image processing steps are built into the chip which attributes to its lower cost.

The quality of fingerprint image captured by optical sensors may get affected by many real world factors. These factors may include stray light from another source or surface contamination such as fingerprint impression left behind by a previous user. Other factors that affect image quality are oil, dirt, condensation or ice and any scratches on the surface.

Compared to other scanner types, it is relatively easy to deceive optical scanners with impostor fingerprints by presenting a convincing picture of a fingerprint. Optical scanner manufacturers have thus introduced a variety of techniques that can validate a live finger. One such technique is electro-optical imaging that enhances optical sensors and improves its resistance to deception. In electro-optical imaging, a voltage is placed across a light-emitting polymer film. When the individual places his finger on the sensor, the fingerprint ridges present a ground to the polymer surface that creates a small current and generates light. This produces a high contrast image as the valleys of the fingerprint remain dark and the polymer is directly linked with the optical detector.

Capacitive scanning is another common way of capturing fingerprint images that utilizes electrical current to sense the image instead of light. The capacitive scanning process uses an array of capacitor plates to capture the image of fingerprint. Similar to optical scanners, this process too generates an image of the ridges and valleys that compose fingerprints. Human skin is conductive enough and is able to provide capacitive coupling in combination with an individual capacitive element on the array. The physical ridges of the fingerprint are closer to the capacitor plates and have a higher capacitance whereas the valleys of the fingerprint i.e. the subdermal layer have a lower capacitance. A small voltage can be applied by some capacitive sensors to enhance the signal that results in an image with better contrast. The capacitive sensor is able to measure the smallest differences in conductivity that are caused by the presence of ridges.

An electrostatic discharge such as shock can interfere with this type of scanner but it is not affected by ambient lighting. Moreover, capacitive based fingerprint scanners can resist contamination issues better than some optical scanners. It is also quite difficult to fool this scanner using a high-quality fingerprint photograph rather than an actual finger. But capacitive scanners can be deceived by using the mould of an individual’s fingertip.

The sensor type is optical in an optical scanner as the name suggests. In optical scanners, the sensor surface does not require any special treatment or maintenance. Optical sensors are very durable in nature. They are scratch-resistant and the glass plate is made of an unbreakable material that is as hard as quartz. Optical scanners are also resistant to shock, extreme weather and ESD. They are designed to perform well in high traffic as well as rough or outdoor environments. It has a large imaging area that results in a larger as well as a higher resolution image. The manufacturing cost is lower in optical scanners. It also has a long life and almost zero maintenance.

A capacitive scanner has a semiconductor or chip type of sensor. The coatings on the surface of capacitive scanners are uneven and wear out over time. This results in a degraded performance and also shortens the lifetime of the product. Capacitive sensors usually need some kind of a surface treatment such as ESD and other protective coatings. These sensors are not as durable as optical sensors and can be damaged by electrostatic discharge. Repeated handling and everyday exposure can corrode the surface easily. Moreover, silicon chips are thin and inherently fragile and prone to be damaged by hard external impact and scratches. A capacitive scanner usually has a smaller imaging area as compared to an optical scanner which results in a smaller image size with lower resolution. It is expensive to produce surface coatings of consistent quality. There can also be other expenses such as replacement, downtime and maintenance costs.

Optical sensors can be fooled by latent prints i.e. the print left behind when a real finger touches the sensor plate. Latent prints are usually produced by sweat, skin debris or other sebaceous excretions that cover up the palmar surface of the fingertips. If a latent print is on the glass platen of the optical sensor and light is directed on it, this print can fool the optical scanner. This happens because the light that is directed on the latent print gets optically scattered and the sensor detects it as a fingerprint image.

Capacitive sensors can be spoofed by using gelatine based soft artificial fingers. This material can mimic the physical characteristics of the skin and hence are able to deceive capacitive sensors.

Multispectral imaging is an optical sensor that has been introduced to reduce the vulnerability of fingerprint sensors to spoof attacks. The speciality of multispectral sensors is that it can capture the features of the tissue that lie below the skin surface as well as the usual features on the finger surface. The features under the skin surface are able to provide a second representation of the pattern on the fingerprint surface. This enables the MSI sensor to collect good quality fingerprint images under a variety of conditions.

In cases where the surface features are worn out or the sampling conditions are adverse, the MSI sensor is still able to capture a fingerprint image. The MSI sensor is strongly able to distinguish between a live finger and other soft materials. The tissue features under the skin surface provide a great amount of information to the MSI sensor about the material that is being imaged and thus makes it less vulnerable to fingerprint sensor attacks.

We have discussed the three types of fingerprint sensors – optical, capacitive and multispectral imaging sensors and their strengths and weaknesses. Multispectral imaging is a fingerprinting breakthrough that is able to provide strong protection against fingerprint spoofing techniques.