中文摘要：

针对当前图像隐写方案存在阶梯效应,使其不可感知能力差,且其信息隐藏容量小（≤50%）等不足,设计了最优像素调整耦合基因算法的高容量图像隐写术。基于HDWT（Hara discerte wavelet transform）机制,构造隐藏信息长度计算模型,找出图像分块的频域表示,以改善隐写鲁棒性;根据载体图像与隐写图像之间的绝对误差,设计适应度函数,借助基因算法,获取最优映射函数,将秘密信息嵌入到HDWT系数中;并设计最优像素变换方案,降低载秘图像与载体图像之间的嵌入误差,显著增大隐写容量;再设计其提取机制,获取信息图像;以PSNR（peak signal to ratio）构建反馈机制,优化提取质量。仿真结果显示,与其他隐写机制相比,该算法具备更大的隐写容量和更强的不可感知性能;拥有更高的检测精度,可有效区分载体与隐写图像特征值。

英文摘要：

In order to sovle these defects such as small capacity of intimation hiding and poor imperceptible performance in current image steganography mechanism, this paper proposed the high capacity image steganography based on optimal pixel adjustment and genetic algorithm. It used HDWT mechanism tofind the frequency domain representation of blocks for improving the robustness of steganography. Then it obtained the optimal mapping function by introducing the genetic algorithm to embed the sceret message into HDWT coeffeints, designed the optimal pixel adjustment algorithm to reduce the erroe between cover image and stego-image and increase the hiding capavity. Finally, it got the resrotation image by the extraction mechaonism. And the simualtion results show that ： comparision with current image steganography mechansim, this algorithm has larger hiding capacity and greater imperceptible performance.