Sub-Saharan African Journal of Medicine

ORIGINAL ARTICLE
Year
: 2016  |  Volume : 3  |  Issue : 1  |  Page : 45--52

Facial biometrics using Akinlolu-Raji image-processing algorithm and anthropological facts which prove that Kebbi and Zamfara Hausas are Hausa Bakwai


Adelaja Abdulazeez Akinlolu 
 Department of Anatomy, Faculty of Basic Medical Sciences, University of Ilorin, Ilorin, Kwara State, Nigeria

Correspondence Address:
Adelaja Abdulazeez Akinlolu
Department of Anatomy, Faculty of Basic Medical Sciences, University of Ilorin, Ilorin, Kwara State
Nigeria

Abstract

Background: Anthropology studies human biology and culture. This study aims to develop a novel image-processing algorithm which can be used to compute cephalometric measurements for face recognition of individuals of any ethnic group or tribe and to produce forensic facial database of Hausas. Materials and Methods: Three hundred Hausas of Kebbi State (150 males and 150 females, aged 18-36 years) were selected as subjects for the study with informed consents and when established as Hausas by parents and grandparents. Height, body weight, and cephalometric parameters (evaluated on three-dimensional facial photographs) were measured on subjects. The novel Akinlolu-Raji image-processing algorithm was developed using modified row method of computer programming. Facial width, total face height, short forehead height, long forehead height, upper face height, nasal bridge length, nose height, morphological face height, and lower face height computed from readings of Akinlolu-Raji image-processing algorithm were analyzed using z-test (P ? 0.05) of 2010 Microsoft Excel statistical software. Results: Biological examination of the anthropological history of Hausas proved that Hausas of Kebbi and Zamfara belong to the Hausa Bakwai States. Comparative statistical analyzes of facial measurements showed sexual dimorphism (P > 0.05) with nonsignificant higher averages in Hausa males compared to females. Hausa males have the leptoprosopic face type while Hausa females have the hyperleptoprosopic face type based on classifications of face types from facial indices. Conclusions: Hausas of Kebbi and Zamfara belong to the Hausa Bakwai States. The novel Akinlolu-Raji image-processing algorithm can be used to compute biometrics for face recognition. Furthermore, sexual dimorphism exists between Hausa males and females.



How to cite this article:
Akinlolu AA. Facial biometrics using Akinlolu-Raji image-processing algorithm and anthropological facts which prove that Kebbi and Zamfara Hausas are Hausa Bakwai.Sub-Saharan Afr J Med 2016;3:45-52


How to cite this URL:
Akinlolu AA. Facial biometrics using Akinlolu-Raji image-processing algorithm and anthropological facts which prove that Kebbi and Zamfara Hausas are Hausa Bakwai. Sub-Saharan Afr J Med [serial online] 2016 [cited 2024 Mar 29 ];3:45-52
Available from: https://www.ssajm.org/text.asp?2016/3/1/45/176320


Full Text

 Introduction



Anthropology studies human biology and culture. [1],[2] In addition, forensic anthropology deals with the establishment of human identity using biometrics and face recognition technology. [3] To provide further information on the anthropological history of Hausas and the roles of craniofacial cephalometry in forensic science and to pioneer the use of face recognition technology in Nigeria, this study aims to review and biologically examine the anthropological history of Hausas, develop a novel image-processing algorithm which can be used to compute cephalometric measurements for face recognition of individuals of any ethnic group or tribe, and provide a forensic facial database of Hausas of Kebbi State. The biological determination of ancestral origins of subjects was not carried out in this study; hence, presented data are preliminary.

 Materials and Methods



Review of Anthropological History of the Hausas

The Hausas were claimed in some traditions to have an ancestral origin named Bayajidda, who fathered three sons, namely Biram (through his first wife, Magira), Karbagari or Karap-da-Gari (through Bagwariya, the slave-maid given to him by Magajiya-the Queen of Daura), and Bawo (through his second wife, Magajiya). Bawo fathered six sons that together with Biram founded the "legitimate seven Hausa Bakwai States" (Daura, Kano, Rano, Zaria, Gobir, Katsina, and Biram). [4],[5] The children of Karbagari were claimed to have founded Zamfara, Kebbi, Nupe, Gwari, Yauri, Yoruba, and Kwararrafa States [4],[5],[6] usually referred to as the seven "illegitimate Banza Bakwai Hausa States," based on the social status of Karbagari's mother as a slave-maid. [4],[5],[6] Alternative or modified versions of Bayajidda legend account are known in Kebbi and Zamfara, but the Bayajidda legend account is unknown in the remaining five "Banza Bakwai Hausa States." [4]

Pilot Study

The pilot study was conducted to determine the reliability of the novel image-processing algorithm using 40 Yorubas (20 males and 20 females), aged 18-23 years, who were undergraduate students of Osun State School of Health Technology, Ilesa and Osun State University, Okuku Campus, as control subjects. Informed consents of subjects were obtained in accordance with ethical guidelines of the Helsinki Declaration of 1975 as revised in 2000. Results of cephalometric parameters generated from manual measurements with the aid of the Vernier caliper and those generated from the measurements read by the image-processing algorithm in control subjects were statistically compared.

Selection of Subjects and Determination of Sample Size in the Main Study

Letters of approval for conduct of the study were received from managements of Kebbi State University of Science and Technology, Aliero; Adamu Augie College of Education, Argungu; Kebbi State School of Nursing and Midwifery, Birnin Kebbi; and School of Health Technology, Jega, from where undergraduate students (150 males and 150 females, aged 18-36 years) were locally selected as subjects for the study using the purposive technique or judgment sampling method. [7],[8],[9],[10] Subjects were selected for the study only when established via distributed questionnaire as Hausas of Kebbi State by parents and grandparents. Informed consents were obtained from subjects in accordance with ethical guidelines of the Helsinki Declaration of 1975 as revised in 2000.

Data Collection and Evaluated Facial Cephalometric Parameters

Data on height and bodyweight, parents and grandparents ethnic origin, local government area, state of origin, and three-dimensional (3D) facial photographs were obtained from each subject. Photographs of subjects were taken with 3D SONY Cyber-shot DSC-HX7V camera (Sony Electronics Incorporated, San Diego, USA) using modified procedures for standardized photography. [11],[12] Facial parameters (in mm) were computed from readings of the Akinlolu-Raji image-processing algorithm on facial photographs of each subject. Height (m) of subjects ranged from 1.6 to 1.9 in males and 1.3 to 1.8 in females while the range of bodyweight in kilograms was 45-85 in males and 43-70 in females.

Distances of the facial width (zygion to zygion), total face height (trichion to gnathion), short forehead height (trichion to glabella), long forehead height (trichion to nasion), upper face height (trichion to subnasale), morphological face height (nasion to gnathion), nasal bridge length (nasion to pronasale), nose height (nasion to subnasale), and lower face height (subnasale to gnathion) were computed in this study [Figure 1].{Figure 1}

Development of the Akinlolu-Raji Image-processing Algorithm for Face Recognition

Digital image-processing employs computer algorithms to process images for recognition by an electronic or computer medium. It permits the application of wider range of algorithms to input data but avoids the build-up of noise and signal distortion in image-processing. [13],[14],[15] Input data generated from images are processed into deciding features that represent the scenes in the image. [13],[14],[15] The novel Akinlolu-Raji image-processing algorithm for forensic face recognition was developed using the modified programming principle of row method. [14],[15],[16] In the row method, each picture element (pixel) given by a number or three-set of numbers called grayscale depending on the color and texture of the image portion being represented was considered column by column along a row until all the rows were covered.

The grayscale of each cell was confirmed to represent the color of the marked points previously set as the threshold grayscale. The coordinates of any detected point were noted and recorded. [14],[15],[16] Since some of the detected points were not at same horizontal or vertical levels, the Pythagoras theorem was used to calculate the pixel distance before converting to actual distance using the pixels of the reference points and their computed distances as read by the novel image-processing algorithm. [14],[15],[16]

For example, computed total face height (trichion to gnathion distance) by the novel image-processing algorithm was converted to actual life size or distance as follows:

Manually computed distance between selected two reference points on a scaled graph sheet: 300 mmComputed distance between the selected two reference points by Akinlolu-Raji image-processing algorithm on 3D facial image: 273 mmComputed total face height (trichion gnathion distance) using Akinlolu-Raji image-processing algorithm on 3D facial image: 131.94 mmConversion of computed total face height distance to life size: 131.94 × 300/273 mm = 131.94 × 1.099 mm = 145 mm.

Statistical Analyses

In the pilot study, computed values of facial parameters in mm (mean ± standard deviation [SD]) from readings of the Akinlolu-Raji image-processing algorithm were statistically compared with Vernier caliper measurements of same facial parameters using the t-test of the Statistical Package for the Social Science software Statistics 23. The alpha value for test of significance was set at P B ≤ 0.05. The Bonferroni correction (P B) method was employed to reduce the chances of obtaining false-positive results (type I errors) declaring wrong significant difference when no significant difference exists. [17],[18],[19]

In the main study, facial parameters computed from readings of the Akinlolu-Raji image-processing algorithm were statistically analyzed using the 2010 Microsoft Excel Statistical software of personal computer manufactured by TOSHIBA Incorporation. The two-sample z-test method (used when the sample size is >30) was employed for statistical significance comparisons of computed means of facial parameters in mm (mean ± SD) between Hausa males and females. The alpha value for test of significance was set at P ≤ 0.05.

 Results



Biometric Measurements of the Face in the Pilot study

In male control subjects, pairwise statistical analyses of results showed nonsignificant higher mean values (P B > 0.05) of long forehead height, total face height, and upper face height, but statistically nonsignificant lower mean values (P B > 0.05) of morphological face height, nose height, and lower face height in computed Vernier caliper values compared to those computed from readings of the Akinlolu-Raji image-processing algorithm [Table 1]. In female control subjects, analyses of results showed statistically nonsignificant higher mean values (P B > 0.05) of long forehead height, total face height, and morphological face height, but statistically nonsignificant lower mean values (P B > 0.05) of nose height, lower face height, and upper face height in Vernier caliper values compared to those computed from readings of the Akinlolu-Raji image-processing algorithm [Table 1].{Table 1}

Biometric Measurements of the Face in the Main Study

Statistical analyses of measurements of anteromedial aspects of the face (mean ± SD in mm) showed nonsignificant higher mean values (P > 0.05) of total face height, short forehead height, long forehead height, upper face height, morphological face height, nasal bridge length, and nose height in Hausa males compared to females. There was statistically nonsignificant lower mean value (P > 0.05) of lower face height in Hausa males compared to females [Table 2].{Table 2}

Three-section Facial Profiles and Facial Indices of Hausa Subjects

Computations showed higher percentages of long forehead height and nose height, but lower percentage of lower face height in Hausa males compared to females [Table 2]. The schematic presentation of the three-section facial profiles of Hausas derived from results presented in [Table 2] is described in [Figure 2]. The facial index (FI) was lower in Hausa males compared to females [Table 2].{Figure 2}

 Discussion



Anthropology studies the historical past and present as well as the specific features that define us as humans. [1],[2] The previous review of the anthropological history of ethnic groups of Nigeria noted migrations, conquests and/or assimilations of indigenous populations by founders from different ethnic origins and re-definitions of ethnic groups across the nation. [4],[5],[6],[20],[21],[22],[23] In addition, anthropometric data computed using ancestry informative markers and phenotype-based race/ethnicity information usually disagreed. [24] Furthermore, analyses of anthropometric data obtained from different studies on members of same "ethnic or racial" group produced conflicting results or data, perhaps due to regional variations. [25]

Therefore, to provide definitive anthropometric data, ancestry informative markers should be employed for precise characterization of individuals and/or collective biological ancestry. [24],[25],[26],[27],[28],[29] In the absence of biological determination of ancestral origins of subjects, anthropometric studies become preliminary, and subjects should be selected from local populations under same environmental or epigenetic influence. In this study, local representative subjects were selected from Hausas of Kebbi State.

The divisions of old Hausa States into Hausa and Banza and the Hausa Bakwai States [2],[4] become difficult to establish as the Afro-Asiatic language of Hausa is spoken in "Hausa Bakwai States" and some "Banza Bakwai States" (Kebbi and Zamfara) as distinct from Benue-Congo languages spoken in the remaining five "Banza Bakwai Hausa" States (Nupe, Gwari, Yauri, Yoruba, and Kwararrafa). [2],[4] In addition, the divisions of old Hausa States into "legitimate" (Hausa Bakwai) and "illegitimate" (Banza Bakwai) on the basis of biological sons of Bayajidda or Bawo bred through either the "wife" or "slave-maid" as previously reported [4] and traditionally believed are invalid with respect to principles of genetics and heredity.

The embryological conception of a new child or zygote (male-XY or female-XX) results from the union of the sperm cell of the father (XY) and the oocyte of the mother (XX). [30],[31] The father releases Y-chromosome while the mother releases X-chromosome in case of the conception of a male child (XY). Similarly, the father releases X-chromosome while the mother releases X-chromosome in case of the conception of a female child (XX). [30],[31] Clearly, the scientific principles of genetics and heredity established biological children as true and legitimate children, irrespective of the social or ethnic status of both or either of the parents. [30],[31] Therefore, Kebbi and Zamfara (founded by Karbagari's children), belong to the Hausa Bakwai States in the same way that Daura, Kano, Rano, Zaria, Gobir, Katsina (founded by Bawo's children), and Biram (founded by Biram) belong to the Hausa Bakwai States.

One-dimensional anthropometry (1D) identifies soft-tissue landmarks over which calipers or measuring tapes are placed for reading distances between landmarks. [8],[32],[33] 1D is disadvantaged by excessive time consumption [12] and possible distortion of soft tissue by its equipment, which may introduce errors, [10] errors of identifications, and readings of anthropometric distances between operators and limited shape information. [32],[33] Digital anthropometry could be two-dimensional (2D) or 3D and are employed in face recognition systems for computing biometric parameters.

The face recognition system (2D or 3D) employs computerized algorithms for face recognition and is the most widely used way of identification or authentication of identity in civil and criminal investigations for forensic analyses and face detection purposes. [34],[35] 2D facial recognition system uses anthropometric equipment such as 2D cameras and is limited by physical appearance changes, changes in lighting intensity, aging, pose and inability to provide structural information of surface curvature, and geodesic distances about the face. [34],[35],[36]

The 3D facial recognition system, in contrast, gives complete and real information of shapes, texture and color, represents shapes or landmarks by set coordinates, provides faster method of data acquisition, and gives more accurate data. It shows a high level of reliability and is more robust to face variations due to different factors. Its algorithm is compatible with variations in illumination conditions during image acquisition [34],[35] and is applicable to both the 2D and 3D face recognition systems. 3D facial recognition system uses 3D digital image technology devices such as surveillance videos, cameras, and scanners with 3D anthropometry for computing biometric parameters. [37],[38] Therefore, 3D anthropometry has potentials in growth assessment studies, clinical analyses, quantification of facial morphology, assessment of facial deformity, anaplastology, genotypic-phenotypic studies of syndromes, and forensic investigations. [37],[38]

Biometric measurements of the face in the present study were conducted in two phases, the pilot study and the main study. In the pilot study, pairwise comparative statistical analyses of computed values of cephalometric parameters between Vernier caliper (1D anthropometry) and Akinlolu-Raji image-processing algorithm (3D anthropometry) measurements in male and female control subjects showed no significant differences (P B > 0.05) in all measured parameters: Total face height, long forehead height, upper face height, morphological face height, nose height, and lower face height [Table 1]. The facial parameters computed from readings of the Akinlolu-Raji image-processing algorithm compared well with measurements obtained using the Vernier caliper. Hence, the image-processing algorithm was confirmed as reliable for further readings of cephalometric measurements in the main study.

In the main study, statistical analyses of 3D anthropometric measurements of anteromedial aspects of the face in Hausas showed statistically nonsignificant higher mean values (P > 0.05) of total face height, short forehead height, long forehead height, upper face height, morphological face height, nasal bridge length, and nose height in males compared to females [Table 2]. There was statistically nonsignificant lower mean value (P > 0.05) of lower face height in males compared to females [Table 2]. The results implied that Hausa males differ from females in facial features (sexual dimorphism) with higher mean values of facial parameters in males than in females.

The results are in conformity with established anatomical principle that females have smaller crania with shorter facial features than males. [39] In addition, the results are in agreement with those of previous 1D anthropometric studies which reported sexual dimorphism with males having higher mean values of facial parameters than females in Binis of Edo State, aged 16-35 years; [40] Igbos of Southeast Nigeria, aged 18-69 years; [41] and Yorubas and Hausas resident in Kano, Kano State, aged 17-25 years. [42]

Computations showed higher percentages of long forehead height (39.3% in males and 35.4% in females) and nose height (26.9% in males and 26.2% in females), but lower percentage of lower face height (33.8% in males and 38.5% in females) in Hausa males compared to females [Table 2] and [Figure 2]. The results implied sexual dimorphism in computed three-section facial profiles between Hausa males and females.

In addition, the three-section facial profiles of Hausa females (long forehead height: 35.4%, nose height: 26.2%, and lower face height: 38.5%) were similar to those of African American females, aged 18-30 years (long forehead height: 37.5% females, nose height: 26.1%, and lower face height: 36.4%) obtained from 1D anthropometry. [25] The percentage proportion of nose height was significantly shorter compared to percentage proportions of long forehead height and lower face height in Hausa females. This result is in agreement with the significant shorter percentage proportion of nose height when compared to percentage proportions of long forehead height and lower face height in African American women, aged 18-30 years [25] [Figure 3].{Figure 3}

The face is classified, based on the FI or prosopic index as hypereuroprosopic or very short broad face (FI < 79.9), europrosopic or short broad face (FI of 80-84.9), mesoprosopic or medium round face (FI of 85-89.9), leptoprosopic or long narrow face (FI of 90-94.9), and hyperleptoprosopic or very long narrow face (FI > 95). [43],[44],[45] The facial indices in Hausas (92.6 in males and 96.6 in females) [Table 2] examined in the present study implied that Hausa males have the leptoprosopic face type while Hausa females have the hyperleptoprosopic face type.

The leptoprosopic face type observed in Hausa males is in agreement with previous 1D studies which reported the leptoprosopic face type in Tangales (FI: 92.1 in males and 92.6 in females) and Tera males (FI: 94.1) of Gombe State. [44] The hyperleptoprosopic face type observed in Hausa females in the present study is in agreement with previous 1D studies which reported hyperleptoprosopic face type in Northeastern Nigerians resident in Maiduguri, Borno State (FI: 99.39 in males and 97.54 in females), aged 19-35 years [43] and Fulanis (FI: 95.2 in males and 100.8 in females) and Tera females (FI: 100.4) of Gombe State, aged 18-40 years. [44] The leptoprosopic and hyperleptoprosopic face types observed in Hausas is in disagreement with previous 1D studies which reported dominant mesoprosopic face type in Malays, aged 19-30 years [45] and Indians, aged 18-22 years (FI: 87.19 in males and 86.75 in females), [46] possibly due to ethnic and regional variations.

 Conclusion



In the present study, biological examination of the anthropological history of Hausas proved that Hausas of Kebbi and Zamfara belong to the Hausa Bakwai States. In addition, based on analyses of data and interpretations of results of facial measurements in control subjects and Hausas of Kebbi State, the novel Akinlolu-Raji image-processing algorithm can be employed for computing anthropometric or forensic measurements on 2D and 3D images. Data computed from readings of the Akinlolu-Raji image-processing algorithm can be converted to actual or life sizes as obtained in 1D measurements. Furthermore, results of facial measurements established sexual dimorphism between Hausa males and females. The biological determination of ancestral origins of subjects shall be carried out in future studies to provide definitive and representative anthropometric data of Nigerian ethnic groups. This will help to determine the true nature of the heterogeneity and ethnic diversity of the Nigerian population.

Declaration of Patient Consent

The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Acknowledgments

I acknowledge the following:

The professional contribution of Professor Raji, Abdulganiy Olayinka, a computer programming expert of the Department of Agricultural and Environmental Engineering of the Faculty of Technology, University of Ibadan, Ibadan, to the development of the Akinlolu-Raji image-processing Algorithm employed for computation of facial cephalometric parameters in this studyThe approval for the conduct of the study and support of the students, staff members and managements of Osun State School of Health Technology, Ilesa, Osun State; Osun State University, Osogbo, Osun State; Kebbi State University of Science and Technology, Aliero, Kebbi State; Adamu Augie College of Education, Argungu, Kebbi State; Kebbi State School of Nursing and Midwifery, Birnin Kebbi, Kebbi State; and the School of Health Technology, Jega, Kebbi State, Nigeria, from where the subjects for the pilot study and the main study were selectedThe approval for the conduct of the study and support of the management of the University of Ilorin, Ilorin, for granting me Staff Development Award, Professor C. N. B. Tagoe and Dr. M. S. Ajao.

Financial Support and Sponsorship

Nil.

Conflicts of Interest

There are no conflicts of interest.

References

1Larsen CS, editor. A Companion to Biological Anthropology. 1 st ed. United Kingdom: Spencer, John Wiley & Sons; 2010. p. 14-38.
2Jurmain R, Kilgore L, Trevathan W. Essentials of Physical Anthropology. 8 th ed. USA: Wadsworth, Cengage Learning; 2011. p. 8-20.
3Kosa F. Application and role of anthropological research in the practice of forensic medicine. Acta Biol Szeged 2000;44:179-88.
4Shillington K. Encyclopedia of African History. A-G. Vol. 1. USA, New York: Fitzroy Dearborn, an Imprint of the Taylor and Francis Group; 2005.
5Lange D. The Bayajidda legend and Hausa history. In: Bruder E, Parfitt T, editors. Africa Zion: Studies in Black Judaism. UK: Cambridge Scholars Publishing; 2012. p. 138-74.
6Shokpeka SA. The relevance of the culture of origin to nation building in Nigeria. J Soc Sci 2009;21:185-9.
7Ted P. Purposive sampling. In: Lisa MG, editor. The Sage Encyclopedia of Qualitative Research Methods. Vol. 1, 2. Thousand Oaks, California, USA: Sage Publications Incorporated; 2008. p. 697-8.
8Akinlolu AA, Akinola BO, Hussein AT. Position of the ear in relation to facial midline landmarks in Nigerians. Internet J Hum Anat 2011;1:1-7. Available from: http://www.ispub.com/IJHA/2/1/10497. [Last accessed on 2015 Oct 10].
9Memon S, Fida M, Shaikh A. Comparison of different craniofacial patterns with pharyngeal widths. J Coll Physicians Surg Pak 2012;22:302-6.
10Bacchetti P, Deeks SG, McCune JM. Breaking free of sample size dogma to perform innovative translational research. Sci Transl Med 2011;3:87:1-4. [Last accessed on 2015 Oct 10].
11Kearney B. Variations of the External Ear in an Australian Population for the Purposes of Identification. A Thesis Submitted to the University of Adelaide in Partial Fulfillment of the Requirements for the Degree of Bachelor of Science, University of Adelaide; 2003. Available from: http://www.eleceng.adelaide.edu.au/personal/dabbott/tamanshud/kearney_oct2003.pdf. [Last accessed on 2015 Oct 10].
12Oghenemavwe EL, Fawehinmi BH, Daenwin TL. A software tool for facial analysis. Res J Appl Sci Eng Technol 2012;4:551-6.
13Pratt WK. Digital Image Processing: PIKS Inside. 3 rd ed. United Kingdom: John Wiley and Sons; 2001.
14Raji AO, Fagboun AA, Dania MK. An Approach to Detecting Defects in Food Products. Proceedings of the First International Conference and 22 nd Annual Conference of the Nigerian Society of Agricultural Engineers, Ibadan, Nigeria; 2000. p. 36-9.
15Ewemoje TA, Raji AO. Evaluation of the leaf area index of cowpea using empirical and image processing methods (Vigna unguiculata). J Eng Res 2011;16:42-52.
16Raji AO. Discrete Element Modeling of the Deformation of Bulk Agricultural Particulates. Ph.D. Thesis, University of Newcastle-upon-Tyne; 1999. Available from: http://www.theses.ncl.ac.uk/dspace/bitstream/10443/871/1/Raji99.pdf. [Last accessed on 2015 Oct 10].
17Wright PS. Adjusted P values for simultaneous inference. Biometrics 1992;48:1005-13.
18David RA, Dennis JS, Thomas AW. Modern Business Statistics with Microsoft Excel. 5 th ed. USA: Cengage Learning; 2015. p. 574.
19International Business Machines Corporation. The Calculation of Bonferroni-adjusted P values; 2015. Available from: http://www.1.ibm.com/support/docview.wss?uid=swg21476685. [Last accessed on 2015 Oct 10].
20Biobaku SO. The Origin of the Yoruba. (Humanities Monograph Series, 1). Lagos: University of Lagos; 1955.
21Johnson S, Johnson O. The History of the Yorubas: From the Earliest Times to the Beginning of the British Protectorate. London: Lowe and Brydone Printers Limited; 1960.
22Gatawa MM. The formation of migrant communities in Nigeria: A study of Gobirawa settlements, Niger State, Nigeria. Eur Sci J 2014;10:86-100.
23Ireyefoju JO, Ireyefoju FE. Ife oracle in Itsekiri social system of Nigeria. Int Lett Soc Humanist Sci 2014;21:62-8.
24Mersha TB, Abebe T. Self-reported race/ethnicity in the age of genomic research: Its potential impact on understanding health disparities. Hum Genomics 2015;9:1.
25Porter PJ, Olson KL. Anthropometric facial analysis of the African American woman. Arch Facial Plast Surg 2001;3:191-7.
26Gonder MK, Holly MM, Floyd AR, Sarah AT. Whole-mt DNA genome sequence analysis of ancient African lineages. Mol Biol Evol 2007;24:757-68.
27Torres JB, Doura MB, Keita SO, Kittles RA. Y chromosome lineages in men of West African descent. PLoS One 2012;7:1.
28Haber M, Gauguier D, Youhanna S, Patterson N, Moorjani P, Laura RB, et al. Genome-wide diversity in the levant reveals recent structuring by culture. PLoS Genet 2013;9:2.
29Luca P, Stephan S, Deepti G, Petr D, Aylwyn S, Yuan C, et al. Tracing the route of modern humans out of Africa by using 225 human genome sequences from Ethiopians and Egyptians. Am J Hum Genet 2015;96:986-91.
30Larsen WJ. Human Embryology. 3 rd ed. Pennsylvania, USA: Churchill Livingstone; 2001.
31Moore KL, Persaud TV. The Developing Human: Clinically Oriented Embryology. 7 th ed. USA: Elsevier; 2003.
32Farkas LG. Anthropometry of the Head and Face. 2 nd ed. USA: Williams and Wilkins, Raven Press, Lippincolt, Baltimore; 1994. p. 3-56.
33Sforza C, de Menezes M. Three - Dimensional face morphometry. Dent Press J Orthod 2010;15:13-5.
34Kukula EP, Elliott SJ, Waupotitsch R, Pesenti B. Effects of Illumination Changes on the Performance of Geometric Face Vision. 38th Annual International Carnahan Conference. Security Technology; 2004. p. 331-7. Available from: http://www.researchgate.net/publication/4130198_Effects_of_illumination_changes_on_the_performance_of_Geometrix_FaceVision_3D_FRS. [Last accessed on 2015 Oct 10].
35Gupta S, Mia KM, Alan CB. Anthropometric 3-D face recognition. Int J Comput Vision 2010;90:331-49. [Doi: 10.1007/ s11263-0100360].
36Moreton R, Morley J. Investigation into the use of photoanthropometry in facial image comparison. Forensic Sci Int 2011;212:231-7.
37Ghoddousi H, Edler R, Haers P, Wertheim D, Greenhill D. Comparison of three methods of facial measurement. Int J Oral Maxillofac Surg 2007;36:250-8.
38Jayaratne YS, Zwahlen RA. Application of digital anthropometry for craniofacial assessment. Craniomaxillofac Trauma Reconstr 2014;7:101-7.
39Moore KL, Agur AMR, Dalley AF. Clinically Oriented Anatomy. 5 th ed. USA: Williams and Wilkins, Lippincott; Baltimore; 2006. p. 887-912.
40Omotoso DR, Oludiran OO, Sakpa CL. Nasofacial anthropometry of adult Bini tribe in Nigeria. Afr J Biomed Res 2011;14:219-21.
41Olotu J, Eroje A, Oladipo G, Edibamode E. Anthropometric study of the facial and nasal length of adult Igbo Ethnic Group in Nigeria. Internet J Biol Anthropol 2009;2:2.
42Anas IY, Saleh MS. Anthropometric comparison of nasal indices between Hausa and Yoruba Ethnic Groups in Nigeria. J Sci Res Rep 2014;3:437-44.
43Raji JM, Garba SH, Numan AI, Waziri MA, Maina MB. Morphological evaluation of head and face shapes in a North-Eastern Nigerian population. Aust J Basic Appl Sci 2010;4:3338-41.
44Maina MB, Mahdi GG, Kalayi GG. Craniofacial forms among three dominant ethnic groups of Gombe State, Nigeria. Int J Morphol 2012;30:211-6.
45Yesmin T, Thwin SS, Urmi SA, Wai MM, Zaini PU, Azwan K. A study of facial index among Malay population. J Anthropol 2014; Article ID 726974:1-4. Available on: http://dx.doi.org/10.1155/2014/726974. [Last accessed on 2015 Oct 10].
46Shetti RV, Pai SR, Sneha GK, Gupta C, Chethan P, Soumya. Study of prosopic (facial) index of Indian and Malaysian students. Int J Morphol 2011;29:1018-21.