The fingerprint image quality is crucial for the performance of a fingerprint recognition system. Many factors may influence the fingerprint image quality, such as, the type of sensor used to capture the image (optical, capacitive), rough fingertips (manual worker, elder people, allergic skin), fingertip condition (wet, dry), image resolution (500 dpi,250 dpi), poor contact of the finger with the sensor, presence of noise, latent images (traces from the previous user), etc.
With the advances of the technology, the quality of the sensor has been improved significantly, however the optical scanners continue performing better than the other, specially for the case of poor quality fingers.
optical sensor
capacitive sensor
thermal sweeping
eletric
The problem of wet or dry fingerprint is very common, as it depends much of environmental condition (temperature, humidity) and contact of the finger with stuffs.
Dry fingerprint are caused normally by low temperatures or after wiping just washed hands. The image obtained from dry fingerprint presents interrupted ridges, as it is shown in the figure below. In order to correct this problem, the user may moisturize their fingerprint simply by passing the finger on the forehead before applying it to the optic window. In the case of wet fingerprint it is recommended to wipe the fingers.
normal fingertip condition (good quality)
dry fingertip (medium quality)
wet fingertip (medium quality)
A low quality of fingerprint can also obtained by absence or poor defined ridges. Aging is a important factor related to the quality of the fingerprint, elder people have worse fingerprint than the younger. Certain occupations, such as construction and farming, are also known to damage fingerprint details over time as well as the constant contact of finger with chemical products.
low quality fingerprint images
Another important factor influencing the quality is the finger pressure during the capture, slight pressure can produce light contrast, with the same effect as dry fingerprint. By the other hand heavy pressure can produce smudged fingerprint images. The image resolution may also influence to the quality, above 500 dpi are considered enough to extract the feature information with good quality, however resolution below 250 dpi may compromise the quality of the feature extraction.
During the enrollment it is expected to capture the image of the best quality as possible, for this reason it is important to quantify the quality of the fingerprint image. if the value of the quality is too low, the image can be discarded, and ask to try to capture another sample. To compute the whole image quality the image are divided in small blocks and computed the quality of each block, for that purpose are used the contrast, the ridge orientation consistency and coherence, ridge frequency, etc. Using the quality one can classify the regions as well defined, recoverable and unrecoverable.