A lossy JPEG2000-based data hiding method for scalable 3D terrain visualization

The data needed for 3D terrain visualization consists, essentially, of a texture image and its corresponding digital elevation model (DEM). A blind data hiding method is proposed for the synchronous unification of this disparate data whereby the lossless discrete wavelet transformed (DWTed) DEM is embedded in the tier-1 coded quantized and DWTed Y component of the texture image from the lossy JPEG2000 pipeline. The multiresolution nature of wavelets provides us the scalability that can cater for the diversity of client capacities in terms of computing, memory and network resources in today's network environment. The results have been interesting and for a bitrate as low as 0.012 bit per pixel (bpp), a satisfactory visualization was realized. We compare the obtained results with those of a previous method that interrupt the lossless JPEG2000 codec immediately after the DWT step and embeds lossless DWTed DEM in the reversibly DWTed Y component of texture. The proposed method proved to be more effective in the sense that for the same bitrate one observed lesser quality loss for respective resolutions.