International Journal of Research and Reviews in Information Sciences (IJRRIS) Vol. 2, No. 1, March 2012, ISSN: 2046-6439 © Science Academy Publisher, United Kingdom www.sciacademypublisher.com 155 Age Range Estimation from Human Face Images Using Face Triangle Formation Hiranmoy Roy 1 , Debotosh Bhattacharjee 2 , Mita Nasipuri 2 , and Dipak Kumar Basu 2 1 Department of Information Technology, RCC Institute of Information Technology, Kolkata 700015, India 2 Department of Computer Science and Engineering, Jadavpur University, Kolkata 700032, India Email: hiranmoy_roy17@yahoo.co.in, debotosh@indiatimes.com, mita_nasipuri@gmail.com, dipakkbasu@gmail.com Abstract – With the advancement in technology, one thing that concerns the whole world and especially in the developing countries is the tremendous increase in the population. With such a rapid rate of increase, it is becoming difficult to recognize each and every individual because we have to maintain copies either in digital or hard copy format of every individual at different time periods of his life. Sometimes database has the required information of that particular individual, but it’s of no use as it is now obsolete. With age a person’s facial features changes and it becomes difficult to identify a person given an image of his at two different ages. This paper discusses a novel mechanism by which two images at different time periods of the individual life can be compared so that it can be ascertained both images are of same individual. This is done by extracting feature points of a face from which a face triangle is formed. If the ratio of areas of the triangles of both images is within a specified range then we can say both images are of same person. A range of threshold values is proposed adhering to the ratios of the areas between various age groups which can be matched with to determine whether a particular image is of the same person or not. At the same time if it is known that the images are of same person then age range can also be estimated. Experimental results show that face recognition and age range estimation both may be effectively performed and which performs with low computational effort. Keywords – Face Triangle, Age Estimation, Face Recognition, Feature extraction, Eye localization, Eyebrow detection 1. Introduction Face recognition is an important field of biometrics which is of great use in our day to day life. Be it the traditional uses in identification documents such as passports, driver’s licenses, voter ID, etc., or its uses i n recent years, where, face images are being increasingly used as additional means of authentication in applications such as credit/debit cards and in places of high security. But with age progression the facial features changes and the database needs to be updated regularly of the changes which is a tedious task. So we need to address the issue of facial aging and come up with a mechanism that identifies a person in spite of the aging. 1.1. Techniques used in Face Recognition Techniques used in face recognition can be broadly categorized into three, as per followings: i) Traditional, which includes methods like, Eigenface or principal component analysis (PCA), fisherface or linear discriminate analysis (LDA) etc. These techniques [1, 2] extract facial features from an image and then using them perform search in the face database for images with matching features. Other algorithms [3,4] normalize a gallery of face images and then compress the face data, only saving the data in the image that is useful for face detection. A probe image is then compared with the face data. ii) 3-D: this technique uses 3-D sensors to capture information about the shape of a face [5, 6]. This information is then used to identify distinctive features on the surface of a face, such as the contour of the eye sockets, nose, and chin. This technique is robust to changes in lighting and viewing angles. iii) Skin texture analysis: this technique [7, 8] uses the visual details of the skin, as captured in standard digital or scanned images, and turns the unique lines, patterns, and spots apparent in a person’s skin into a mathematical space.. 1.2. Earlier work on Age based Face Recognition A lot of work has been done previously in the field of face recognition starting with Gibson’s ecological approach towards perception [9] and Thompson’s pioneering work on geometric transformations in the study of morphogenesis [10] that mostly laid the fundamentals for the study of craniofacial growth. They explained human head as a fluid filled spherical water container and performed a hydrostatic analysis on the effects of gravity on craniofacial growth. In human computer interaction, aging effects in human faces has been studied from two main reasons: 1) automatic age estimation for face image classification and 2) automatic age progression for face recognition. Kwon et al. [11] developed a system to classify face images into one of three age groups: infants, young adults and senior adults. They extracted key landmarks from face images and calculated distances between those landmarks. Then ratios of those distances were used to classify face images as that of infants or adults. They also proposed methods for wrinkle detection in predetermined regions in face images to further classify adult images into young adults and senior adults. The first real human age estimation theory was proposed Lanitis et al. [12, 13]. They proposed methods to imitate aging effects on face images.