Biometric Identification via PCA and ICA Based Pattern Recognition

Biometric pattern is the unique signature of people. Biometric verification is a very complicated procedure involving technologies of pattern recognition, signal processing and image processing. In most cases it is necessary to employ artificial intelligence based approaches. This work is focused on intelligent control applications on biometric verification. The actual sensing information is digitized into image matrix files and then data matrices are analyzed using advanced algorithms. Thus, the real patterns of fingerprints will be captured. The information is indicated by plaintexts containing inherent signatures. After transforming plaintext data into ciphertext data, each of three RGB intensity components of the trimulus color system is computed and analyzed individually. Then principal component analysis (PCA) and independent component analysis (ICA) are applied for decision making. Its assumption is evident and very effective that biometric fingerprints can be effectively recognized by utilizing advanced artificial intelligence technologies.