Based on the prediction error expansion algorithm, an improved reversible data hiding algorithm is proposed with the gray image. In order to improve the data embedding rate and reduce the auxiliary data capacity, the algorithm proposed in the paper is used to determine the edge pixels and their adjacent pixels, which can guarantee that the embedded pixel does not have an underflow or overflow problem. Finally, experiments on natural images and medical images show that the data hiding algorithm has better effect on the embedding capacity and image quality.%利用灰度图像的值特征，本文在预测误差扩展算法基础上，提出一种改进的可逆数据隐藏算法。该算法对边界像素结合其邻接像素进行判断，在保证不会产生溢出时嵌入数据，以提高原载体图像的数据嵌入率，并降低辅助数据容量。最后选取自然图像与医学图像进行实验证明，该数据隐藏算法在嵌入容量、图像视觉质量方法均取得较好的效果。
展开▼
