Video Steganography Algorithm Based On The DCT Coefficient Classification

Steganography as one of the important branches of information security,is becoming a Research hotspot.Due to H.264 video is widely used in the world,therefore,based on H.264 video steganaly algorithm research has important theoretical significance and application value.A steganographic algorithm based on DCT coefficient module 3 operation classification is proposed.Aiming at the shortcomings of the required number of vectors such as matrix coding and so on,the concealment and the security are poor,and a new class of steganography is explored.First of all,DCT coefficients are classified,and then the number of categories,the use of the number of categories for steganographic embedded.This idea greatly reduces the number of carriers required,and is well suited for real-time steganography in the absence of a carrier.The steganographic algorithm based on the DCT coefficient module 3 is based on this new idea,that is,the quantization of the DCT coefficient on the 3 to take the remaining number of 0,1 and 2 three categories,the statistics of these three categories of coefficients The number of bits with the remainder of 1 and 2 the number of two types of embedded 2-bit information.By contrasting with other algorithms,the algorithm has very few modifications to the coefficients,and the quality of the video is less than that of other algorithms.A steganalysis algorithm based on DCT coefficient combination classification is proposed.The DCT coefficients are linearly combined using the idea of the EMD(Exploiting Modification Direction)algorithm.That is,the quantization of the discrete cosine transform domain in the eight low-frequency coefficients,the two pairs of groups after a linear combination,mapped into a 5-state 5-bit information bits,statistics,the number of four types of embedded 4-bit information.The experimental results show that the proposed algorithm has more obvious advantages over other algorithms such as peak SNR and embedding rate.