抗统计分析的BPCS隐写算法

位平面复杂度分割（BPCS）隐写在复杂度较高的位面小块嵌入秘密信息，有较好的隐蔽性，但会引起图像位面小块复杂度的统计特性的变化。通过深入分析图像位平面上相邻像素之间的相关性，本文提出了一种改进的BPCS隐写法。它更好地利用人的视觉特性，结合图像的局部特征来嵌入秘密信息。实验结果表明，改进后的算法能够抵御针对原BPCS隐写的分析方法，有较大的嵌入量，并对原图像所作的改变具有更好的不可见性。

BPCS steganography algorithm against statistical analysis

Bit-Plane Complexity Segmentation(BPCS) steganography replaces bit-plane blocks of high complexity with secret data.It provides good imperceptibility.But it would change the statistic characteristics of bit-plane block complexity.Through detailed analysis of the relativity of the consecutive pixels in digital image bit-plane,an improved BPCS steganography is proposed.The algorithm takes full advantages of human visual characteristics and incorporates the local feature of digital image while inserting secret data.Experimental results show that,improved algorithm can resist analysis of original BPCS and provide good capacity.It also has better invisibility for the change of original image.