Fingerprint identification system

Method of comparing fingerprints from a repository file and a Search Candidate file comprising the steps of

• scanning each fingerprint to determine the quality of the fingerprint image,
• scanning each fingerprint to determine the location and quality of each minutia found,
• storing for each fingerprint in said repository file and said Search Candidate file an image state map which identifies:
• a. good quality areas where the image quality is above a first predetermined value which also contain minutiae,
• b. good quality areas where the image quality is above a second predetermined value which also contain no minutiae,
• c. areas where the image quality is below said second predetermined value, or where the image quality is above said first predetermined value but where the quality of minutia found is below a third predetermined value, and
• comparing corresponding fingerprints from said repository file and said Search Candidate file by providing a positive score based on the correspondence between good quality areas containing minutiae and good quality areas containing no minutiae, and a negative score where minutiae are present in a good quality area in one file where no minutiae are present in a corresponding good quality area in the other file print.

Method of comparing fingerprints from a repository file and a Search Candidate file comprising the steps of

• scanning each fingerprint to determine the ridge angle of the fingerprint at a plurality of regularly spaced intervals to create a Ridge Angle Map,
• scanning each fingerprint to determine the quality of the fingerprint image,
• scanning each fingerprint to determine the direction, location and quality of each minutia found,
• storing for each fingerprint in said repository file and said Search Candidate file an image state map which identifies:
• a. good quality areas where the image quality is above a first predetermined value which also contain minutiae,
• b. good quality areas where the image quality is above a second predetermined value which also contain no minutiae,
• c. areas where the image quality is below said second predetermined value, or where the image quality is above said first predetermined value but where the quality of minutia found is below a third predetermined value,
• comparing corresponding fingerprints from said repository file and said Search Candidate file by
• a. aligning each fingerprint to be compared by maximizing the correlation between corresponding Ridge Angle Maps, and
• b. providing a positive score based on the correspondence between good quality areas containing minutiae and good quality areas containing no minutiae, and a negative score where minutiae are present in a good quality area in one file fingerprint where no minutiae are present in a corresponding good quality area in the other file print.

A method of measuring and storing a set of quality parameters as part of a Products File that provides heretofore unused information for improvement in overall fingerprint identification system performance, said information being gathered to provide for superior performance of the Classifier, Searching, and Matching components of system, said information allowing the Search and Matcher components of the system to function dynamically from comparing a Search Candidate ten-print Products File to each prospective File Candidate ten-print Products File and also function in dynamic manner in deciding which pairing of same numbered finger to use between comparison of the Search Candidate's ten-print Products File to a particular File Candidate ten-print Products File, said pairing being those sets of fingers which have high quality and useable feature data in common, so that matching is performed on the basis of best available data between Search Candidate file and any particular repository File Candidate, wherein the quality data measured and store:

• storing minutiae quality indicating confidence, or strength, of determination of being a minutiae feature,
• core and delta quality indicating confidence, or strength, of determination of being a core or delta feature,
• core and delta average local image quality at the location and in the immediate vicinity,
• an image state map which defines both high quality areas, with and without minutiae, and areas of both low image quality and/or of too weak an indication to be declared as containing minutiae,
• a fingerprint summary quality metric for each finger, including
• a. average fingerprint quality of the overall fingerprint image obtained by averaging all smoothed quality matrix pixel points whose value is above a predetermined threshold,
• b. feature area count obtained from the image state or final quality matrix by counting its pixel points that show the presence of high quality/high confidence features of minutiae and non minutiae areas.