brand logo
logo

Digital solutions can greatly aid generation of facial sketches in criminal justice process

29 Jun 2022

  • Researchers note high degree of accuracy when anthropometric measurements used in composite images
  BY Ruwan Laknath Jayakody Owing to the significance of the accuracy level of composite images when it uses anthropometric (of or relating to the scientific study of the measurements and proportions of the human body) indices as measurements, a computerised solution will enhance the current manual facial composite procedure and will benefit the provision of services to the criminal justice process such as to law enforcement officers and legal counsel, and thus ensure that victimised persons have quick access to justice.  These observations were made in a research article on “Forensic art: Two Dimensional (2D) facial composite through image processing techniques in Sri Lanka” which was authored by L. Sivaneasharajah, M.A.S. Perera, P.B. Jayasekara, D.D. Karunaratne, K.D. Sandaruwan and R.N. Rajapakse (attached to the Colombo University’s School of Computing), and J. Perera (attached to the same University’s Medical Faculty’s Forensic Medicine and Toxicology Department), and published in the Sri Lanka Journal of Forensic Medicine, Science and Law 7 (2) in February, 2017. A facial composite is a graphical representation of a human face according to the description of an eyewitness or eyewitnesses. In Sri Lanka, the use of forensic art for the purpose of identifying suspects is done through manual hand drawing in order to obtain a facial composite. However, according to an interview conducted by the researchers with an interviewee from the Criminal Records Division of the Police Department on the manual process of suspect identification through facial composite in Sri Lanka, the failure rate of suspect identification through the manual facial composite sketch was reportedly quite high in 2014 in Sri Lanka (declared to be as high as 92.86%).  This higher failure rate emphasises the necessity of an immediate solution for the facial composite. It is also important to realise that not identifying perpetrators of crime may leave the country in a precarious position as there is a seemingly rising number of crimes in the country. In order to overcome this situation, per P.B. Jayasekara, L. Sivaneasharajah, M.A.S. Perera, J. Perera, D.D. Karunaratne, K.D. Sandaruwan and R.N. Rajapakse’s “Facial index based 2D facial composite process for forensic investigation in Sri Lanka”, an automated, image processing-based, computer software solution has been introduced with 2D facial feature templates by incorporating both medically defined indices and aesthetic aspects. Even though there are numerous international software, per C. Frowd, D. McQuiston-Surrett, S. Anandaciva, C. Ireland and P. Hancock’s “An evaluation of the United States systems for facial composite production”, these software are not the best suited to be adopted to the Sri Lankan context as the facial feature templates which are available in these software may not be relevant for the Sri Lankan population as noted in the aforementioned interview.  Since Sri Lanka is at its infancy level in terms of facial composite, Sivaneasharajah et al. attempted to bridge the gap in order to find out the optimum way to automate this traditional procedure by incorporating the information technologies (image processing) and statistical analysis techniques and its related aspects. Thus, as a first step, facial feature templates need to be constructed by analysing a dataset from the Sri Lankan population.  According to R. Rajapakse, A. Madugalla, I. Amarasinghe, V. Padmathilake, A. Dharmaratne, D. Sandaruwan and M. Vidanapathirana’s “Facial muscle anatomy based approach for forensic facial reconstruction in Sri Lanka”, it was proven that the ability to assess facial feature appearance has been achieved by anthropometric proportion indices and that this has been used for a number of forensic practices.  Hence, when constructing the facial feature templates, Sivaneasharajah et al. focused on two major parameters known as the most occurring indices and commonly available facial feature shapes. In order to find the relative measurements for these parameters, two sub-researches on facial anthropometric indices and shape classification were carried out targeting the local population.  Therefore, in the first phase of the research, the aforementioned sub researches were conducted with 140 undergraduate students (both male and female) aged 20-25 years to find out the measurements of anthropometric proportion indices and the shape classification for the eyes, nose, face, upper vermilion and lower vermilion (the vermilion border/margin/zone is the normally sharp demarcation between the lip and the adjacent normal skin and which represents the change in the epidermis [the outermost layer of skin on the body] from the highly keratinised [the process in which the cytoplasm (the substance surrounding a cell nucleus) in the top layer of skin cells, changes into, or changes so as to contain keratin (a strong natural protein, the main substance that forms hair, nails, horns, feathers, etc.)] external skin to the less keratinised internal skin) for Sri Lankans. Moreover, the entire data set was categorised based on the weight categories according to the Body Mass Index (BMI) values as the BMI can be used to screen different weight categories. The index measurements were measured with the use of the Face Software Development Kit library as it is comparatively more reliable and accurate over manual measuring techniques while the domain expert’s knowledge from the aesthetic and forensic art field were utilised in identifying the facial feature shapes.  Afterwards, by incorporating these parameters (the facial feature template creation process by incorporating anthropometric indices and commonly available shapes involves weight categories, the index measurement of the facial features and commonly available shapes of facial features, and arriving at 2D facial feature templates), 2D facial feature templates for the face, eye, nose and mouth were created targeting Sri Lankan people and eventually these templates were then transferred into the system which can be used for the purpose of composite construction. In addition, this system enables filtering the most possible feature templates based on the eyewitness description given with regard to the age, height, weight and other criteria. Besides, Sivaneasharajah et al. have identified the template positioning of each facial feature by analysing the distance between the facial features and incorporating the standard divine proportions (Y. Jefferson’s “Skeletal types: Key to unravelling the mystery of facial beauty and its biological significance”) which is suitable for the Sri Lankan context. Finally, the face visualisation process will perform iteratively until the eyewitness is satisfied with the ultimate composite image. This solution has been evaluated by Jayasekara et al. with face pool and anthropometric index evaluation techniques where the face pool evaluation technique has obtained 70.19% accuracy, while on the other hand, the anthropometric index evaluation method obtained an average of 84% out of the population who gave a response regarding the matching of the faces.


More News..