Multimodal Biometric Recognition in Feature Level Fusion using Statistical Moment Measure of Color Values

Dhiman Karmakar

Assistant Professor, Dept. of Computer Science, Surendranath College, Kolkata, INDIA.


Nowadays inclusion of more than one biometric trait in human recognition system is a trend to ensure the robustness of the system. This article first incorporates a feature level fusion of face and iris images of personnel and then performs a statistical moment based measure on the RGB color components of these images. A weighted difference between the computed moment vectors is used for classification. This novel approach yields a satisfactory rate of recognition and outperforms unimodal system. The work is dedicated to all the researchers working in the concerned field.

Keywords :Unimodal, Multimodal, PCA, Moment, Fusion, NN.

map1 map2 map3