Human facial image age group classification based on third order four pixel pattern (TOFP)‎ of wavelet image

Abstract EN

The present paper proposes a novel scheme for age group classification based on Third Order Four pixel Pattern (TOFP).

This paper identified TOFP patterns in two forms of diamond pattern which have four pixels i.e., outer diamond and inner diamond patterns in Third Order neighborhood.

The paper derives Grey-Level Co-occurrence Matrix (GLCM) of a Wavelet image based on the values of Outer Diamond Corner Pixels (ODCP) of TOFP and Inner Diamond Corner Pixels (IDCP) of TOFP on wavelet image which is generated from the original image without using the standard method for generating the co-occurrence matrix.

Four GLCM features are extracted from the generated matrix.

Based on these feature values, the age group of the human facial image was categorized.

In this paper, human age is classified into six age groups such as Child: 0-9 years, Adolescent: 10-19 years, Young Adult: 20 – 35 years, Middle-Aged Adults: 36 - 45 years, Senior Adults 46 – 60 years, Senior Citizen: age > 60.

The proposed method is tested on different databases and comparative results are given