Abstract: This paper proposes a reversible watermark algorithm with larger embedding capacity. Through a prediction way, it obtains an approximate value of the existing value, and embeds the watermark by using the differential value relation between the approximate value and the existing value. In extraction process, after extracting hided information, the original image is recovered. Based on the algorithms, another algorithm is proposed by adding information head, which has higher prediction accuracy. Experimental result shows that, compared with other reversible algorithms, the algorithms proposed have fewer modifications of the original maps with larger embedding capacity.

Key words: reversible watermark, large-capacity, precision tolerance, lossless compression

摘要： 提出一种拥有更大水印嵌入量的可逆水印算法。以预测的方法获得一个与现有值近似的值，利用该近似值与现有坐标值的差值进行水印的嵌入。提取时先得到水印信息，再恢复出原始图像。在此基础上，通过添加头信息，提出第2种算法，以提高预测准确度。实验结果表明，与同类可逆算法相比，提出的2种算法拥有更大的水印嵌入量，对原图的改动更少。

关键词: 可逆水印, 大容量, 精度容忍度, 无损压缩

CLC Number: 

• TP301.6