<title>Steganography using wavelet compressed images</title>

Abstract

Internet bandwidth is in high demand, and one way that web sites lower the amount of bandwidth they use is by compressing their site's images. This lowers the amount of bandwidth used, and makes the site load much faster. There are of course many other useful applications for compressed images. Bit Plane Complexity Segmentation (BPCS) digital picture steganography is a technique to hide data inside an image file. BPCS achieves high embedding rates with low distortion based on the theory that noise-like regions in a bit-plane can be replaced with noise-like secret data without discernible loss in image quality. This is possible because the human eye, while very good at distinguishing anomalies in areas of homogenous texture, is bad at distinguishing anomalies in visually complex areas. However, BPCS is not a robust embedding scheme, and any lossy compression usually destroys the data. Wavelet image compression using the Discreet Wavelet Transform (DWT) is the basis of many modern compression schemes. The coefficients generated by certain wavelet transforms have many image-like qualities. These qualities can be exploited to allow BPCS to be performed on the coefficients. The results can then be losslessly encoded, combining the good compression of the DWT with the high embedding rates of BPCS.