LSB replacement is a spatial-domain steganography technique that hides information by replacing the least significant bit of image samples with message bits.
Why it matters
LSB replacement is simple, widely known, and implemented by many practical tools. Its simplicity also makes it a useful baseline for understanding image steganography and steganalysis.
The main weakness is that direct bit replacement introduces statistical artifacts. These artifacts can be exploited by structural steganalysis methods such as Sample Pair Analysis (SPA), which do not require training on an image collection.
Typical tools
Tools associated with this family include:
- OpenStego;
- OpenPuff, for some image modes;
- CryptoStego, for some image modes;
- simple educational or legacy image steganography tools.
Detectability results
In the current StegoWatch comparison, LSB replacement is represented in the uncompressed-image comparison. The results include both SPA and machine-learning-based detection.
The important point is not only that LSB replacement can be detected, but that it can be attacked with structural methods that are less dependent on training data than deep-learning detectors.
Limitations
The available comparison depends on the evaluated image collection, payloads, implementation details, and detectors. Results should not be generalized to every possible image source or implementation.