Region Based Age Classification using Cross Diagonal Centre Symmetric Motif Matrix (CD-CS-MM)
Nara Sreekanth1, Munaga HM Krishna Prasad2
1Nara Sreekanth, Research Scholar, Associate Professor, Department of CSE, BVRIT Hyderabad College of Engineering, Bachupally (Telengana), India.
2Munaga HM Krishna Prasad, Professor, Department of Computer Science and Engineering, University College of Engineering JNTUK, Pithapuram Road, Kakinada (Andhra Pradesh), India.
Manuscript received on 07 March 2019 | Revised Manuscript received on 20 March 2019 | Manuscript published on 30 March 2019 | PP: 976-984 | Volume-8 Issue-5, March 2019 | Retrieval Number: E3172038519/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Age group classification of human beings based on their facial images plays a vital role in many application including security, low and enforcement etc… The extraction of significant features from the facial textures plays a crucial role in age classification. The precise and significant features from facial images can be derived based on local, region or global based methods: out of which local based methods exhibits good results. This paper derived the facial features from local and macro regions. This paper initially divided the facial image into micro regions of size 2×2. The Motifs are derived on each 2×2 grid and the grid is replaced with Motif coded image. This paper divides the image into micro region of size 3×3, where each value represents a Motif index value of a 2×2 grid. This research derives centre symmetric relationship and also measured the cross and diagonal relationship between the Motif codes. This transforms the image into a cross diagonal centre symmetric Motif coded image (CD-CS-MC) image where the code ranging from 0 to 31. The gray level co-occurrence matrix (GLCM) features are derived by deriving a co-occurrence matrix on CD-CSMC and these results a CD-CS-Motif Matrix (CD-CS-MM). The CD-CS-MM is tested on popular facial databases and tested with by dividing the age groups into 4 and 3 levels. The experimental results reveal the efficacy of the proposed method over the other methods.
Keyword: Texture, Motif, Micro Region, Texture, GLCM.
Scope of the Article: Classification