Fingerprint Based Gender Classification using multi- class SVM

Gender identification from fingerprints is an imperative footstep in forensic science in order to deliver fact-finding hints for finding anonymous persons. Fingerprint verification is unquestionably the most honest and satisfactory substantiation till date in the court of law. Due to the massive potential of fingerprints as an effective technique of identification an attempt has been made in the present work to analyze their correlation with gender of an individual. Due to the immense potential of fingerprints as an effective method of identification an attempt has been made in the present work to summarize all the recent techniques related to the field and we have proposed a technique for gender classification which will use some features of finger such as ridge thickness, ridge density to valley thickness ratio (RTVTR) and ridge measurement for gender detection. Proposed methodology uses Multi Class SVM as classifier which overcome the problem of SVM (Binary Classifier).