摘要:
基于色調(diào)處理技術(shù),該文給出了一種有效可行且具有自修復(fù)能力的圖像認(rèn)證算法。首先,基于誤差擴(kuò)散色調(diào)處理技術(shù)將水印圖像4bit色調(diào)量化,井依據(jù)混沌置亂算子,將色調(diào)結(jié)果置亂,然后構(gòu)造平均誤差最小的特征集合C,最后依據(jù)誤差擴(kuò)散數(shù)據(jù)隱藏算法將置亂后水印圖像隱藏于原始圖像中;在認(rèn)證端,從接收到的圖像提取其中所隱藏水印信息并進(jìn)行逆置亂,比較接收到的圖像和反置亂后的隱藏信息,判斷內(nèi)容發(fā)生變化的位置,并依據(jù)所提取的水印信息修復(fù)被篡改圖像。實(shí)驗(yàn)結(jié)果表明,該算法對刪除、替換、篡改等破壞圖像內(nèi)容的惡意操作有精確的檢測和定位,以及自修復(fù)能力。
Abstract:
Based on the image halftone technique, a new effective image authentication algorithm is proposed with self repairing ability. First, the original image is quantized by 4 bit based 011 the error diffusion image halftone, and then the result is permuted by chaos operator, at the same time the halftone color set is constructed with the least quantization error. Next, the permuted image is hidden into the original image based on the error diffusion watermark algorithm. When the watermarked image is received, to verify the authenticity of the received image, the hidden information is extracted out from the received image and permuted inversely. If the received image is similar to the extracted information, the received image is creditable, else that is to say, there exists a local region that is dissimilarity, i.e. the region is tampered, according to corresponding region of extracted information, the tampered region is repaired. Finally, experimental result to demonstrate the effectiveness of the proposed image authentication technique is presented.
下載:
