SUMMARY

See the on-line or hardcopy review (when it appears); we only summarize here.

Hiding messages within a "container" or "cover" leads to study of detecting existence and/or size of an imbed; specific messages or features; restoration of damaged containers; detection evasion; and frustrating messages. The author lays out some Hidden Markov Model math and identifies parallels to a general problem description. Information typically is assumed to be partial, and methods depend on what is available. The author focuses on couple of tasks, and even for them, finds a need for moving from basic to mode advanced models. Author simulation results from previous work come into play offering some experiences and help in choices of how to proceed. Some presumably known information about appropriateness of certain types of testing seems to be reinvented here. Claimed novelty and uniformity are matters a reader will want to attend to, though the author's next work (and possibly his previous conference paper) may be need to satisfy the reader. Minor defects in the paper are sometimes awkward phraseology and inattention to some details even in the definitional stages.