DjVu is a multi-layer raster image file format for digital documents. It was originally developed at AT&T Labs, and is commonly used in book digitization, for example by the Internet Archive.
DjVu documents may include a plain text layer (e.g. from OCR), as well as other data such as a document outline, so the format can serve some of the same purposes as PDF.
Files have a 4-byte preamble. The rest of the file uses IFF format.
Files begin with ASCII characters "
At offset 12 should be a tag indicating the specific file type. For DjVu v3, the possibilities are "
DJVI", and "
There is an extension of DjVu called Secure DjVu. Secure DjVu files begin with "
- DjVu v3 Reference (requires DjVu plug-in)
- DjVu 1999-04-29 (v2) Reference (requires DjVu plug-in)
- Secure DjVu Specification (requires DjVu plug-in)
- The Specifications documents listed above
- The DjVuLibre distributions include some DjVu files.