File formats, compression and Practical Problems
If you actually used to transmit information will use this channel, there are some practical problems. Most image and audio files contain sufficient natural noise at a 3% change to hide. Here we see an ordinary black and white image:
he noise is clearly visible. Here is a change in the LSBs not continue on. So it is quite possible to hide information in this picture.
Many image and audio files provide enough noise to hide data. The first image provides enough material, so that small changes are not noticed. But there are also pictures, information not incorporated good hiding. Images that were generated in the computer, such as Paint with Bush or the Adobe Illustrator or Corel Draw. These areas are great filled with a color evenly. Even if this picture is only slightly changed, one can easily see this as the uniform color noise is superimposed with one. (Actually, this uniform colors hurt my eyes. Why change something artists such areas and overlay it with a texture. This one has then again the possibility to hide the fact data.)
Data formats are a serious problem for anyone who wants to use steganography bit level on a regular basis to hide data. Many image and audio file formats are designed to filter out the noise around to save space. This can be a very special and efficient file format such as GIF happen by, or through very aggressive Kompresionsprogramm, such as JPEG, which therefore does not care whether the picture is the same reconstruction due, as before. Both formats make it hard, just the least significant bit of the image to hide information.
The GIF file format and can be 8-bit standard is a major obstacle, since 8-bit color very 24-bit colors are different. Here is a table of 256 colors used. These selected colors represent the colors in the image on best. The color of each pixel is now an entry in the colormap described through, and even the color of the actual next comes on. The bits have no connection with the intensity of the color of each pixel. So if you changed the least important bit, it does not mean that one slightly by 1% maximum intensity changed, but that one instead of the entry 16 of the color table (light yellow) then entry 17 (dark blue) is replaced. Although they differ only in unimportant bit, but that can produce a major difference in the result.
There are a lot of different solutions to this problem. One way is to use smaller number of colors to one.
.....