摘要:
信息認證是網(wǎng)絡安全的一個重要方面,其目的在于判斷信息的完整性和可信性。該文在對以往的圖像認證技術分析的基礎之上,給出了一種基于感興趣區(qū)域的圖像認證算法。首先將要認證的圖像分割成一定大小的基本圖像塊,然后根據(jù)用戶給出的感興趣區(qū)域,后續(xù)分割每一個基本圖像塊,并提取相應各子塊的簽名信息并保存分割附加碼,最后將簽名信息隱藏于圖像子塊的中頻系數(shù)中。實驗結果證明算法不僅對偶然攻擊具有較強的魯棒性,而且對惡意攻擊具有較高的檢測、定位能力,同時具有較強的自適應能力。
Abstract:
Information authentication technique is an important inspects of network security, and its motive lies in authentication of the integrity and creditability of information. Based on the analysis of the previous image authentication techniques, a new effective image authentication technique is proposed based on the interested region. Firstly, the original image is divided into a certain size preliminary image blocks. Then on the basis of the interested region given by users, the preliminary image block is divided into smaller blocks, at the same time the signature of each image block is extracted. Finally the signature of image blocks is embedded into intermediate frequency coefficient of image blocks. Experimental results are provided to demonstrate the effectiveness, self-adaptability, and validity of the method.