Reconstruction of Fingerprints using Convolutional Autoencoders
Saatwik Bisaria · International Journal for Research in Applied Science and Engineering Technology · 2021
Phones unlock instantly when a finger registered with them is placed on the sensor, but it refuses to recognize the other unregistered finger, that's where the uniqueness of fingerprints becomes practically noticeable.Fingerprints can be used to identify a single person because they are unique to each person and do not change over time.Fingerprints consist of ridges that are elevated lines and grooves that are valleys between these lines.And hence fingerprints are patterns of ridges and furrows that are different for everyone.Ridge patterns are what is imprinted on the surface when your finger touches it.If your fingerprints are taken and printed on paper, they can be used to match fingerprints that you might have left elsewhere. Autoencoders are special types of neural network architectures in which the output is the same as the input. An autoencoder is a regression task where the network is asked to predict its input (in other words, model the identity function). These networks have a tight bottleneck of a few neurons in the middle, forcing them to create effective representations that compress the input into a low-dimensional code that can be used by the decoder to reproduce the original input. Using Autoencoders, we can find a possible way of recreating a fingerprint image with a dataset of already provided fingerprints which can be used as a tool for forensics , biometric investigations , genetics as well as genealogy purposes.