Publication details

Cloud-computing based system for analysis of 3D facial data

Authors

CHALÁS Igor KOHÚT Ľuboš FERKOVÁ Zuzana SOCHOR Jiří

Type Conference abstract
MU Faculty or unit

Faculty of Informatics

Citation
Description Research in the field of forensic biology, anthropology, pathology and many others is highly dependent on the accessibility and/or amount of high quality data available. While working with the biometric data, such as 3D images of human faces we often encounter several law-related and security restrictions, due to sensitivity of the data. This presents several obstacles when accessing the data. We propose a novel system allowing analysis of large dataset of 3D facial images to general public, while maintaining high security standards. The system is realized by connecting The FIDENTIS Analyst software with The FIDENTIS 3D Face Database, performing the bulk of computations on server side. The FIDENTIS Analyst enables a variety of 3D facial morphological analyses designed for forensic purposes, such as 3D facial composite construction, automated landmark localization, face-to-face comparison and analysis of facial morphological variation via batch processing. The FIDENTIS 3D Face Database represents one of the largest-numbered datasets of 3D facial scans with more than 2000 individuals. The system makes data processing available to the user without accessing the data itself, moreover it allows for simplified collaboration among research teams. Thanks to our cloud-computing based system we allow users to utilize the database taking into account the permissions of the user. As most computations are performed on the server side, analysis does not impose special requirements on user’s hardware even for more complex datasets. With this model of computation we are able to provide high security of the process, which is important for 3D facial images. The system offers new way of accessing the biometric data for research in forensic science.