Can facial proportions taken from images be of use for ageing in cases of suspected child pornography? A pilot study

Abstract

The age of the victim plays a crucial role for the legal implications concerning pornography. Judges therefore often call on forensic experts to verify the age of individuals depicted on photographs or videos. However, there is no scientifically established protocol available for forensic practice in such cases. The conventional methods such as the evaluation of secondary sexual characteristics provide unsatisfactory results particularly when the legally relevant ages for child pornography (i.e. 14 and 18 years) are concerned. To overcome these limits, a European research group has explored the applicability of facial proportions as an age indicator on images. In this pilot study, standardized facial images of 353 females and 20 males from four age groups (6, 10, 14 and 18 years) were randomly selected for the metric analysis from a large data set including German, Italian and Lithuanian subjects. In this sample, several indices extracted from the frontal and lateral photographs were closely correlated to their respective indices taken from the living individuals. Furthermore, age-related changes were identified for indices taken from the photographs. The discriminant analysis showed that for the pooled sample, 60.3% of the cases were correctly classified into the respective age group. The percentage of correctly classified cases increased in the respective country samples as follows: 69.9% for Germany, 69.4% for Lithuania and 80.5% for Italy. The present study suggests that the metric assessment of the face may be used for age estimation on images. Nonetheless, more work needs to be done in order to verify the reliability of these findings on a large sample.

Age Estimation in Living Individuals

“…As a whole, growth in height and depth as opposed to growth in breadth predominates in development. Although the assessment of body and facial proportions alone only yields a rough age classification, it should always be included in the overall assessment (Gabriel et al 2011;Cattaneo et al 2012). …”

Juvenile age estimation from facial images

“…In cases involving pornographic images suspected to depict children, it is of great importance to be able to assess if the individual in the image is indeed a child, thus indicating the crime of child sexual abuse (CSA) (3). A more precise age is often required to determine the severity of offender sentencing and may be used in the identification of the victim (4,5).…”

“…Consequently, there has been a rise in the number of cases in which an expert has been called upon to give an age assessment from a photograph of alleged CSA (4,5). Experts may include individuals such as paediatricians, gynaecologists, forensic pathologists or forensic anthropologists (3,4), although this remains a very contentious area with only very recent attempts to provide guidelines for use in age estimation from photographic images (9-11).…”

Proposta de metodologia de análise fotoantropométrica para identificação humana em imagens faciais em norma frontal

“…Considerando a definição cefalométrica "clássica" do ponto Násio (GEORGE, 2007; KOLAR; SALTER, 1997;CATTANEO et al, 2012;ZIMBLER;HAM, 2005) O levantamento da acurácia na determinação dos pontos de referência é inviável nas análises fotoantropométricas, pela inexistência de uma localização que possa ser atribuída como correta ou verdadeira. Neste sentido, buscou-se uma descrição minuciosa para que estes fossem determinados da forma mais precisa possível, ou seja, com reprodutibilidade.…”

“…Ambas as fases foram analisadas por cinco examinadores, os quais marcaram 16 pontos na topografia facial de 18 imagens, aleatoriamente escolhidas de um banco de imagens. Para todas as análises, foi utilizado um software desenvolvido pelo Serviço de Tabela 2 -Descrição anatômica dos pontos cefalométricos entregue aos examinadores de acordo com os autores George (2007), Kolar e Salter (1997), Cattaneo et al (2012) Gráfico 2 -Representação gráfica da dispersão média (considerando dois desvios padrão) das marcações obtidas em ambas as fases, para as distâncias em "x" e "y". A primeira coluna representa a dispersão média da Fase 1, a segunda representa a dispersão média da Fase 2 e a terceira coluna, a dispersão média de ambas sobrepostas.…”