The Pattern That Sets You Apart
Look closely at the tip of any finger and you will see a fine pattern of ridges and valleys swirling across the skin. That pattern, called a fingerprint, forms before birth and stays effectively unchanged for the rest of your life. Even identical twins have different fingerprints. A fingerprint scanner is built around this simple fact: your fingertip carries a signature that no one else has exactly.
Capturing the Ridges
To recognize a finger, the scanner first has to see it. Different scanners do this in different ways. Optical scanners work much like a tiny camera, shining light onto the finger and photographing the ridge pattern. Capacitive scanners, the kind found in many phones, work with electricity instead. The sensor has a fine grid of microscopic electrical points, and when the finger touches it, the ridges actually contact the surface while the valleys leave a tiny gap of air. The scanner can detect this difference electrically, building a map of the ridges without ever using light.
Turning the Pattern into Data
The captured image of the ridges is not stored as a picture. Instead, the device picks out the distinctive details — small features such as where ridges end or split — and converts them into a compact set of numbers. This mathematical description is the "fingerprint template." When you set up the scanner for the first time, the phone records one of these templates for each finger you train and keeps it locked away in protected storage.
Matching at Unlock Time
When you later place a finger on the sensor to unlock, the scanner runs the same process again, building a fresh template from the live touch. The device then compares the new template to the stored one. The two will never be perfectly identical — angle, pressure, and moisture all change the image a little — so the scanner looks for a close enough match by a strict similarity test. If it matches, the phone unlocks. If it does not, it stays locked, and you fall back to a passcode.
A Body as a Key
Fingerprint unlock is one example of biometric identification — using a physical trait of the body as a form of identity. It is the same idea behind facial recognition, just using a different feature: ridges on a fingertip instead of distances between points on a face. Both rely on the same principle that nature gives every person a small set of measurements that are practically unique to them.
Source
This article was written using information from Wikipedia.