NCJ Number
48312
Date Published
1972
Length
5 pages
Annotation
RESULTS OF A 1972 PROJECT SEARCH TECHNOLOGY COMMITTEE STUDY INTO THE FEASIBILITY OF HOLOGRAPHIC ASSISTANCE TO FINGERPRINT IDENTIFICATION ARE REPORTED.
Abstract
ALTHOUGH A NAME SEARCH PROCEDURE IS CAPABLE OF PROVIDING SUCCESSFUL IDENTIFICATION FOR APPROXIMATELY 65 PERCENT OF INCOMING FINGERPRINTS, APPROXIMATELY 80 PERCENT OF IDENTIFICATION BUREAU COSTS GO TO THE SEARCH FOR THE REMAINING 35 PERCENT. THEREFORE, THE MAJOR AVENUE FOR COST REDUCTION LIES IN THE TECHNICAL SEARCH PROCESS. TO TEST THE IDENTIFICATION SUCCESS OF HOLOGRAPHIC PROCEDURES, A TEST FINGERPRINT LIBRARY OF 10,000 FINGERPRINT CARDS WAS ASSEMBLED WHICH HAD A HENRY CLASSIFICATION DISTRIBUTION REPRESENTATIVE OF THAT FOUND IN THE CALIFORNIA MASTER FILES. AN ADDITIONAL 600 CARDS, SOME OF WHICH DID NOT HAVE MATCHES WITH THE TEST LIBRARY, WERE ALSO ASSEMBLED FOR IDENTIFICATION. OF THESE, 100 WERE USED AS A TUNING SET FOR PRETESTING AND ADJUSTMENT OF HOLOGRAPHIC EQUIPMENT, 100 WERE USED FOR AN INITIAL TEST, AND 400 WERE USED FOR A SECONDARY TEST. THREE CORPORATIONS PARTICIPATED IN THE TESTING, EACH USING A HOLOGRAPHIC CONFIGURATION DEVELOPED BY THAT CORPORATION. THE PARTICIPANTS INCLUDED THE KMS TECHNOLOGY CENTER, THE MCDONALD DOUGLAS ELECTRONICS COMPANY (MDEC) AND SPERRY-RAND RESEARCH CENTER (SRRC). ONLY MDEC AND SRRC PARTICIPATED IN THE SECOND TESTING. MULTIPLE RESPONSES WERE ALLOWED FOR EACH CARD (UP TO 10), REFLECTING THE BELIEF THAT HUMAN VISUAL INSPECTION OF CARDS WILL ALWAYS BE NECESSARY AT THE END OF THE PROCESS. DATA WAS RETURNED FROM THE CORPORATIONS FOR ANALYSIS. RESULTS WERE SCORED FOR CORRECT MATCHES WITH TEST LIBRARY, CORRECT DISMISSALS (CORRECTLY IDENTIFIED AS HAVING NO MATCH IN THE LIBRARY), FALSE DISMISSAL (INCORRECTLY IDENTIFIED AS NOT MATCHING LIBRARY), MISMATCH, AND NONPROCESSED. ONLY ONE CARD WAS NONPROCESSED BY ANY COMPANY, INDICATING THAT THE HOLOGRAPHIC PROCESS IS CAPABLE OF PROCESSING THE NORMAL SPREAD OF FINGERPRINT IMPRESSION QUALITY. CORRECT MATCHES FOR THE THREE COMPANIES RANGED BETWEEN 88 AND 96 PERCENT. THESE RESULTS COMPARE FAVORABLY WITH THE APPROXIMATELY 93 PERCENT CORRECT MATCH RATE ESTIMATED FOR A GOOD MANUAL SEARCH, AND IN THE CASE OF MDEC THE 96 PERCENT MATCHING EXCEEDS MANUAL ACCURACY STANDARDS. WHILE ACCURACY IS CRITICAL, FEASIBILITY OF HOLOGRAPHIC ASSISTANCE ALSO DEPENDS UPON PROCESSING TIME, WHICH IN THE CASES STUDIED WAS CONSIDERABLY GREATER THAN THAT REQUIRED BY MANUAL SYSTEMS. PROCESSING TIME HOWEVER, HAS BEEN SUBSTANTIALLY REDUCED BY TWO OF THE CORPORATIONS SINCE THE TIME OF THE EXPERIMENT. CHARTS AND SUPPORTING DATA ARE PROVIDED. (JAP)