Vector-quantization-based reversible information hiding and extraction method

Abstract

The invention, which belongs to the technical field of reversible information hiding, relates to a vector-quantization-based reversible information hiding and extraction method. The method comprises: a sending terminal carries out recompression on an index table after vector quantization of an image and carries out secret information embedding; and a receiving terminal carries out secret information extraction and image and index table recovery. To be specific, the sending terminal divides indexes in the index table that is obtained by vector quantization on the image into seed indexes and remaining indexes and recompresses the remaining indexes in the index table based on correlation among the indexes and an established layered state code book to obtain spatial embedded secret information; and the receiving terminal recovers the seeding indexes for a received code stream, recovers the remaining indexes based on identifier bits and the embedded secret information, and then according to the recovered index table, carries out decoding by using vector quantization to obtain an image. The method has the following advantages: the embedding capacity of secret information can be increased effectively; the compression rate is reduced; and the embedding rate is enhanced.

Description

technical field [0001] The invention belongs to the field of information hiding, and in particular relates to a reversible information hiding and extraction method based on vector quantization. Background technique [0002] With the development of Internet technology, digital images are widely disseminated in the network as an important medium. At the same time, in order to ensure the security and confidentiality of some important information, image-based information hiding technology has been paid more and more attention. Information hiding is to embed secret information into an ordinary digital carrier without affecting the subjective quality of the digital carrier, and it is not easy to be noticed by the observer. However, in previous information hiding schemes, the digital carrier cannot be recovered after extracting the secret information, which is not applicable to some fields with specific requirements. Therefore, based on this requirement, reversible information hid...

AI Technical Summary

Problems solved by technology

[0004] At present, the reversible information hiding schemes based on the vector quantization compression domain can be divided into the following categories: one is to recompress the index table and then embed secret information through the search-order-coding theory. The image with strong correlation has obvious effect, but the effect is not obvious for complex image; the second is to recompress the index table through the local adaptive coding theory (Locally-adaptive-coding), that is to say, for the same index with The same codeword encoding, and then embed secret information in the obtained space, which can save time and improve efficiency, but this method also requires high correlation between indexes; the third is through the adjacent state codebook mapping theory ( Adjoining state-codebook mapping) recompresses the index table, generates two state codebooks by using adjacent indexes, and then combines the two codebooks as the state codebook of the current index to compress the current index, and then obtains the space embedding secret information and many more

[0005] The above three schemes all propose a theory to recompress the index table after vector quantization, and then embed secret information. Although a certain capacity of secret information can be embedded, the above methods do not use the correlation between indexes. Very sufficient, the recompression of the index table is not very thorough, resulting in a small embedded capacity

Images