PAPERS
To return to the "Published Papers" main page, click here.
THE SEVENTH INTERNATIONAL CONFERENCE ON FORENSIC COMPUTER SCIENCE - ICoFCS 2012
Print ISBN 978-85-65069-08-3 - Online ISBN 978-85-65069-06-9, pages 12-19
DOI: 10.5769/C2012002 and http://dx.doi.org/10.5769/C2012002
An adaptive sampling strategy for automatic detection of child pornographic videos
By Pedro Monteiro da Silva Eleuterio, and Mateus de Castro Polastro
To download this paper, click here.
ABSTRACT
Automatic detection of child pornographic files is a desirable task for law enforcement, especially at crime scenes, where pedophiles can be arrested immediately. The NuDetective Forensic Tool was developed to quickly identify these illegal files, even at crime scenes. To detect child pornography videos, an approach was prior developed, based on extraction and sampling frames of video files. The previous strategy established a fixed number of video frames to be sampled, independently of the video duration. This work proposes a new adaptive sampling approach, considering the video duration, with the objective to increase the detection rate and/or reduce the runtime. Several experiments were performed and the results proved that this new approach is more appropriate to be used in the detection of child pornographic videos at crime scenes, with detection rates around 87% and a reduction
KEYWORDS
child pornography, nudity detection, video frame extraction, computer forensics, crime scene analysis.