Open AccessReversible data embedding for vector quantization compressed images using search‐order coding and index parity matching
Author(s) -
Qin Chuan,
Chang ChinChen,
Horng Gwoboa,
Huang YingHsuan,
Chen YenChang
Publication year - 2015
Publication title -
security and communication networks
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.446
H-Index - 43
eISSN - 1939-0122
pISSN - 1939-0114
DOI - 10.1002/sec.1046
Subject(s) - vector quantization , embedding , information hiding , computer science , steganography , algorithm , coding (social sciences) , quantization (signal processing) , index (typography) , pattern recognition (psychology) , artificial intelligence , mathematics , statistics , world wide web
Embedding secret data in vector quantization (VQ) compressed images with reversibility has been studied extensively in recent years. However, to date, the reported methods have not achieved satisfactory performances of hiding capacity and image compression ratio simultaneously. In this paper, we propose a reversible embedding method based on search‐order coding (SOC) and index parity matching that can hide secret data into the compressed VQ index, that is, SOC index. If the parity of the candidate SOC index matches the current embedding bit and the error caused by SOC encoding is smaller than a pre‐determined threshold, the length of the stego SOC index after embedding is significantly shorter than the original VQ index. On the receiver side, the embedded secret bits can be easily extracted by checking the parity of stego SOC indices, and all original VQ indices can be recovered correctly. Experimental results demonstrate that our method can achieve greater hiding capacity than the recently reported methods for the same image decompression quality. Copyright © 2014 John Wiley & Sons, Ltd.