Research On Color Image Reversible Data Hiding

Today in the information explosion, information security has been paid more andmore attention.In the aspect of image, information hiding technology has been the hotresearch spot, as a branch of image information hiding, reversible data hiding is usedvery importantly in many important fields in the real life, such as medical image,military and legal evidence etc. This paper mainly studies the reversible data hidingbased on color image. In the aspect of reversible data hiding, difference expansionembedding method is the most classic one, which, by Tian, is first applied to thereversible data hiding and has received very good effect. In recent years, many scholarsmade a number of improvements on the basis of Tian algorithm.This paper mainly carried out two aspects of research work based on the Tianclassical algorithm: that is how to improve the secret data embedding capacity of vectorcolor image and how to improve the quality of the vector color image after embedingsecret data. Looked after a large number of domestic and international literature onreversible data hiding, after analysis of the Tian classical algorithm and some of itsimproved algorithms, This article Mainly do work on the basis following three items:①This paper analyzes and summarizes the classic algorithm which Tian proposed,as well as some improvement extend algorithm some scholars proposed based onclassical algorithms in recent years.②This paper proposes a reversible data hiding algorithm for the twice dataembedding. In some of the existing algorithms, we found that a considerable of the dataembedded locations are rather special, pixel values will not be changed after embeddingthe data, where can be embed data again. The experimental result shows that: thisalgorithm not only improves the embedding capacity of data, but also the quality ofvector image can also maintain a good after embedding data. It is better than someexisting reversible data hiding algorithms.③Existing reversible data embedded algorithms mostly improve embeddingcapacity and image quality, which are contradictory.This article will introduce anembedding extent factor Ψ which is used to measure how good of the extent of an algorithm which take balance between embedding capacity and quality of the vectorcolor image. In other words, it is used to measure the improvement level of embeddingcapacity when the magnitude of a loss in quality is the same. Ψ is a ratio of the growthof embedded rate and the PSNR change of vectors image.