JBIG2 encoder

[Dingo]

Googlebooks is able to store several scanned images of a book in a single pdf, with high efficient compression (uses JBIG2 I think)

looking for JBIG2 encoder, I've found this

http://imperialviolet.org/jbig2.html

no binaries available. any idea about how to build from sources? this encoder may solve my problems in storing scanned images books in a pdf high compressed, so reduce filesize as max as possible