Analysis of fingerprint images for recognition of persons - determining incidence of different greyscale values in pixels and extracting characteristic parameters of living finger

摘要

The method involves detecting a fingerprint image comprising pixels of different greyscale values using a fingerprint sensor. Further fingerprint images are detected and classified in a predetermined class. A region of pixels is sub-divided into intervals, and the number of pixels whose greyscale values lie in the same interval is determined for each image. A parameter is derived from these numbers and used to classify the fingerprint images into respective classes. Histograms and sum histograms of the incidences of different greyscale values may be determined and a characteristic parameter of a living finger extracted.