Identical Twins Face Recognition System Based on Time Series Motif Discovery

In the last few decades the birth rate of twins has been increased, thus the need for an accurate biometric system to precisely determine the identity of a person who has an identical twin became of great interest especially in criminal cases. Recent researches has showed the performance of automatic face recognition technology fails drastically in case of identical twin siblings compared with other unrelated persons. In this paper, we propose a new technique for identifying identical twins using their time series which is unique feature for image. The characteristic time series dimensionality reduced by using Piecewise Linear Approximation(PLA) method. A recognition and matching experiment conducted on an identical twins of high-similarity correlation coefficient score 0.96 and the methodology was able to distinguish between them even with fewer number of features; 8-features