Automatic Dermatoglyphics Multiple Intelligence Test based on palmprint Analysis using Convolution Neural Network

Abstract

Palmprints play a very crucial role in uniquely identifying a person; but it can only be able to
derive the innate potentials, multiple intelligence, and personality as well from human palmprints.
Currently, the Dermatoglyphics Multiple Intelligences (DMIT) is used to generate reports by indicating
various traits and potentials exhibited by a person. This research paper has been proposed to automate
the manual work carried out by an expert. Image processing techniques and convolutional neural
networks are used in the development of the proposed system. Image processing techniques are used to
perform various operations on palmprint images to remove the noise. Then the noise-free images are
sent to Convolutional Neural network to correctly identify the palmprint type out of the 5 classes.
Further, the correct classification is used to categorize his/her personality, learning type, acquiring
ability and the individual perception rate. The results obtained in this research work are encouraging as
the accuracy remains at the acceptable level by considering that the sample size used in the process is
inadequate.