Research On Information Hiding Algorithm Based On Game Texture Compression Format

Today,the game industry has developed into an emerging industry with great potential and broad application prospects.Facing the huge market space,an increasing number of enterprises want to get high profits from it.Plagiarism,copying and piracy phenomenon o ften occur.Among them,game texture image data is often stolen illegally,which leads to the interests of the game developers are damaged.Game texture images are different from ordinary digital images,which need to be compressed and stored by texture compression algorithms.The direct application of conventional information hiding algorithms will lead to the problems that the secret information cannot be extracted or the image quality is damaged.In this paper,three kinds of information hiding algorithms based on PowerVR Texture Compression(PVRTC)are proposed,aiming at copyright protection of texture images in games and combining with related algorithms of information hiding.The main research work of this paper is as follows:(1)When the information hiding algorithm based on hamming code is applied to PVRTC data,there will be a large distortion problem in the uncompressed game texture image.This paper proposes a PVRTC information hiding algorithm based on improved hamming+1.This algorithm combines the characteristics of PVRTC data,and realizes information hiding by preprocessing the modulation information in PVRTC data and a new modification strategy.The algorithm includes the following steps:First,the game texture image is compressed by PVRTC 1(4bpp)algorithm to obtain compressed data;Second,select the modulation information of "01 " and " 10" in the compressed data and group them;Third,the Hamming+1 algorithm is combined with a new modification strategy to complete the secret information embedding operation for each group of modulation information;Finally,the carrier data embedded with secret information is obtained.Experimental results show that the game texture image distortion is smaller,the image quality is higher,and the embedding capacity is higher.(2)When the reversible information hiding algorithm based on Pixel Value Ordering(PVO)is applied to PVRTC data,the prediction error is not concentrated on"-1","0" and "1",which leads to the problem of low embedding capacity.This paper proposes a PVRTC reversible information hiding algorithm based on pixel value ranking and prediction error statistics.The algorithm using the ideas of pixel values sort,and combining the prediction error statistics to realize reversible information hiding.The algorithm includes the following steps:First,the game texture image is compressed by PVRTC1(4bpp)algorithm to obtain compressed data;Second,read the color value information of the compressed data to form two low-resolution images,and use the checkerboard overlay algorithm to divide it into blocks;Third,the pixel values in each block are sorted and the prediction errors are counted,and the secret information is embedded in each block according to the statistical prediction error;Finally,the carrier data embedded with secret information is obtained.Experimental results show that the algorithm has a high embedding capacity and is superior to other algorithms.(3)When the weighted matrix-based information hiding algorithm is applied to PVRTC data,the embedding capacity is low and data irreversible due to the fact that the weighted matrix only performs the embedding operation once,so this paper proposes a weighted matrix-based PVRTC large-capacity reversible information hiding algorithm.The algorithm generates multiple weighted matrix according to the initial weighted matrix and complete secret information embedded in many times,and combining with index file to realize reversible information hiding.The algorithm includes the following steps:First,the game texture image is compressed by PVRTC1(4bpp)algorithm to obtain compressed data;Second,read the color value information of the compressed data to form two low-resolution images and divide them into blocks;Third,the secret information is embedded in each block by using multiple weighted matrices;Finally,the carrier data embedded with secret information is obtained.Experimental results show that the algorithm has higher embedding capacity and efficiency.