Research On Novel Steganography Algorithm With High Capacity

As an essential mechanism in secured communication, the steganography technology has become one of the hotspots that raise the concern from both academic and national defense industry fields. Normally steganography methods can be divided into two categories:spatial and transformation domain; and each can be further categorized into different types such as low awareness, high capacity, high robustness, etc., according to the emphasis of different evaluation indices.Through the recognition of and improvement to existing algorithms, this research tries to discover some new steganography idea in high level of hiding capacity. The whole research can be divided roughly to spatial part and transformation part, coming up with 1 or 2 new methods. In transformation part, based on widespread JPEG format, discrete cosine transformation (DCT) is studied for secret embedding. The general idea is embedding through DCT of extended blocks, and the choosing of proper quantization table for good effect and evaluation indices.Spatial domain gets its eyes on the encoding and grouping techniques for the secret, and complexity of pixel blocks, each spawning a hiding method. Simple LSB way is studied in the first one, and a generalized LSB method is formed after conclusion and expanding, with a rise in bit-per-pixel; the complexity is metered in certain way for the second one, then decision on whether to embed secret in one block is made via threshold setting.