植基於空間域之彩色浮水印

中文摘要在科技技術一日千里的今天，隨著許多科技產品之軟硬體價格的平民化，以及網際網路之普及，彩色數位影像作品隨處可得，以致數位版權難以分辨。而浮水印技術就是眾所周知，用來解決數位版權之爭的一個解決方案。但大多數的浮水印技術都是使用灰階影像來嵌入二元浮水印，在色彩繽紛的多媒體世界裡，這樣的顏色數實在是太少了。因此本論文提出一個可以使用較少計算量之演算法來嵌入彩色浮水印於彩色影像之中，且無需原始影像資料之佐證，即可萃取出所嵌入之彩色浮水印技術。首先將欲藏入之彩色浮水印利用統計的方法取出色彩最多的前24色，以此24色為彩色浮水印之代表色，並且將此彩色浮水印之表代色以一個亂數種子所產生的亂數序列來進行攪亂，而此亂數種子可視為一個秘密金鑰。接著我們以被嵌入之彩色影像在RGB色彩空間中之B-Channel為對象，利用相鄰像素間俱有相仿灰階值之關係，以2x2區塊像素間大小順序不同的排列組合，來代表24種不同的色彩，並且可藉由一個參數d來加大像素間彼此的距離，以增加其浮水印之強韌性。在取出浮水印時將從受測影像中取出2x2區塊像素間大小順序排列組合與之對照，此時不須參考原始影像資料即可取出浮水印。實驗證明我們所提出的方法比已經提出之方法優良。

關鍵字

彩色浮水印、2x2區塊、空間域、頻率域、離散餘弦、離散小波

並列摘要

Abstract Today, technology is at a tremendous pace, the software and hardware prices are acceptable, and the popularization of Internet, the color images can be obtained easily. It is a difficult problem for affirming copyrights, and digital watermark is a well known technology to judge the copyright dispute. But the most watermark technology embed the binary watermark in gray-scale image, in the multimedia world, such solution is insufficient. This thesis proposes a method can use less calculation to embed the color watermark in color image, and does not need the original image to extract the embedded color watermark. First, this study counts the most 24 colors of color watermark that we want to embed, and use these colors as the representative of color watermark. Then uses a pseudo random number generator to scramble the represented color watermark, and saves the random seed as the secret key. This proposed method embeds the color watermark in B-channel of the RGB color space. We can utilize adjoin pixel relation of similar gray-scale level, with the different permutation order according to pixel values of 2x2 block, to represent 24 kinds of different color, and can emphasize the distance among pixels with a parameter d to increase the robustness of the watermark. While extracting the embedded watermark, only need scan the permutation order of 2x2 block for the disputant image, and need not compare the original image. The experiment results prove the method proposed by this thesis is finer than methods have been proposed in the literatures.

並列關鍵字

color watermark, 2x2 block, spatial domain, frequency domain, discrete cosine transform, discrete wavelet transform

參考文獻

[11] 蔡曜州, “建構於空間-頻率域多重浮水印技術之數位影像著作權保護與認證,” 中原大學電機工程研究所碩士論文, 2003。

[14] 倪明輝, “以影像統計資訊為基礎之強韌性空間浮水印技術,” 中原大學電機工程研究所碩士論文, 2003。

[5] S.C. Pei and C. M. Cheng, “Pallete-based color image watermarking using neural network training and repeated LSB insertion,” in Proceedings of the 13th IPPR Conference On Computer Vision, Graphic and Image processing. Vol. 1, pp. 1-8, 2000.

[2] D. P. Mukherjee, S. Maitra, and S. T. Acton, “Spatial Domain Digital Watermarking of Multimedia Objects for Buyer Authentication,” IEEE Transactions on Multimedia, Vol. 6, No. 1, pp. 1-15, 2004.

[3] C. C. Thien and J. C. Lin, “A simple and high-hiding capacity method for hiding digit-by-digit data in images based on modulus function,” Pattern Recognition 36, pp. 2875-2881, 2003.

國際替代計量

植基於空間域之彩色浮水印

未授權

主題瀏覽