Robust and Large Hiding Capacity Steganography using Spread Spectrum and Discrete Cosine Transform

Most information hiding methods present in theliterature have one general principle in common: they hide thesecret message within an insignificant component, such asnoise, of the cover or carrier. Most of these methods do notpay attention to the statistical properties of the noisecomponents and change the statistical profile of the coversignificantly hence leading to the simplicity of breaking thesystem by a passive attacker. In this paper an adaptivesteganography algorithm is proposed that encrypts the secretmessage and spreads it in highly noisy image regions of acarrier image using spread spectrum and discrete cosinetransform methods. A modified message embedding method isalso introduced that is based on comparative blocks, transformand Bit Plane Complexity Segmentation (BPCS) embeddingtechniques to increase the data hiding capacity. Traditionaltechniques have limited data hiding capacities of up to 18% ofthe data amount of a carrier image. The integration of thesetechniques increased the hiding capacity to approximately62%.