(University of Management and Technolog, 2016) Khalid, Mashal; Naeem, Faizana; Awais, Muhammad; Abdulrehman Asghar
The present research aims to verify the utility and reliability of hand and footprint for the identification of an individual. We have work on research papers, for which foot and hand prints of 283 volunteers were taken from all over the Punjab (Pakistan) including 142 males and 141 females aged 18-65. Different foot parameters i.e. Ridges of different areas, the length of toes, toe lengths ratios and Heel-Ball index and hand measurements i.e. Hand finger lengths, hand breadth and ratios of hands were obtained. Various methods have been used for this purpose i.e. ROC curve analysis and different algorithms for gender classification (Naïve Bayes, j48, Random Tree, Random Forest and REP Tree). The accuracy rates for gender determination through classification algorithms were high as compare to gender determination through ROC curve analysis. The study was conducted on Stature estimation with the help of foot parameters by correlation analysis method with reasonable accuracy.
A desktop based software was developed which provides basic functionality related to gender classification and height estimation with the help of hand and foot parameters. In which one can identify the gender and stature through different combinations of hand or foot parameters required by the system. Data mining techniques like classification algorithms for gender determination and regression analysis for stature estimation with more than 80% accuracy