LILUG General Meeting - Tuesday, 9 January, 2007
From Lilug
Meeting Summary
Speakers
- Bart Mallio - Steganography
- Bart started by giving a brief history of steganograpy and passing secret messages. This part of the discussion included references from ancient times straight through World War 2. His presentation then moved on to talk about how much easier it is to encode binary information in binary data.
- He discussed how to evaluate a steganographic protocol by determining the protocol's capacity, security and robustness. There are three basic steganographic methods: injection, substitution and generation.
- Bart then gave an example of a simple steganographic protocol which flipped bits in a bitmap graphic. He explained that this is done by flipping the least significant bit of a series of pixels in the bitmap. He walked through an example of encoding a message.
- Bart then discussed some of the weaknesses associated with using steganography as a communication medium. A group-wide discussion then erupted which was about how discoverable the existence of steganographic data would be in various files and various situations. Bart's talk then transitioned into talking about real tools that can be used for injecting steganographic data in to graphics, namely steghide. He showed us a cover image and then showed us the image after unencrypted text had been hidden inside of it.
- The future of steganography was then covered by Bart where he mentioned what steganographers are looking to do with their data next. This included only adding information to "high energy" areas of sounds and images as well as other advances in the field.
- He went on to give a list of anecdotal steganographic uses that was both informative and humorous. This transitioned in to stories of people trying to find a steganographic image in the wild that is the Internet.
- The talk then switched gears to talk about how to attack an image which contains steganographic data. This involved discussing how certain file formats work and how one might determine that an image contains data. The lower the noise level that is introduced by the injected data, the harder it will be to determine that there is data there at all.
- Bart then discussed host forensics which basically boils down to trying to determine if someone has been employing steganography by looking for steganographic software on their computer. This would also include finding cover images that were left on a steganographer's computer.
- Bart then fielded questions from a bunch of the members on a bunch of different topics. Discussions evolved from subliminal messages and advertising through jpeg steganography and using clips from video to further hide steganographic data.
Discussions/Announcements
Possible Room Change
- Matt Newhall asked if people liked the room that we were currently in. The group's members were largely indifferent but we did discuss all of the things that we like about our current meeting room like the projector and the outlets on the tables.
Raffles
- Computer w/ CRT: Ant
- Ambient Findability from O'Reilly: Walter Rosenblatt
- Enterprise DMZs from O'Reilly: John Oliver
- Linux in a Windows World: Louis G
- Enemy at the Water Cooler: Walter Rosenblatt
- Thicker Than Blood by: Matt Newhall: Mark Drago
Contributed Notes on Meeting From Others
- Video of talk not made public at the request of speaker.