Proposal / Submission Type
Peer Reviewed Paper
Location
Henderson Welcome Center
Start Date
15-5-2017 3:15 PM
Abstract
A common task in digital forensics investigations is to identify known contraband images. This is typically achieved by calculating a cryptographic digest, using hashing algorithms such as SHA256, for each image on a given media, comparing individual digests with a database of known contraband. However, the large capacities of modern storage media, and increased time pressure on forensics examiners, necessitates that more efficient processing mechanisms be developed. This work describes a technique for creating signatures for images of the PNG format which only requires a tiny fraction of the file to effectively distinguish between a large number of images. Highly distinct, and compact, such analysis lays the foundation for future work in fast forensics filtering using subsets of evidential data.
Scholarly Commons Citation
McKeown, Sean; Russell, Gordon; and Leimich, Petra, "Fast Filtering of Known PNG Files Using Early File Features" (2017). Annual ADFSL Conference on Digital Forensics, Security and Law. 1.
https://commons.erau.edu/adfsl/2017/papers/1
Full Resolution File
Fast Filtering of Known PNG Files Using Early File Features
Henderson Welcome Center
A common task in digital forensics investigations is to identify known contraband images. This is typically achieved by calculating a cryptographic digest, using hashing algorithms such as SHA256, for each image on a given media, comparing individual digests with a database of known contraband. However, the large capacities of modern storage media, and increased time pressure on forensics examiners, necessitates that more efficient processing mechanisms be developed. This work describes a technique for creating signatures for images of the PNG format which only requires a tiny fraction of the file to effectively distinguish between a large number of images. Highly distinct, and compact, such analysis lays the foundation for future work in fast forensics filtering using subsets of evidential data.
Comments
View the agenda session- Afternoon Session 2 - Image Forensics