基于位平面分布熵的图像检索算法

Image Retrieval Based on Bit-plane Distribution Entropy

该文针对利用颜色直方图检索时存在的问题,提出了一种基于位平面分布熵的图像检索算法。首先将图像分解为8个位平面,然后采用对表征图像结构特征有意义的4个位平面的信息熵组成的熵矢量来多层次地对图像特征进行描述。为了避免图像中像素灰度值的微小变化对位平面的影响,又提出了采用位平面的灰度码表示方法。同时,考虑到位平面间的相关性,设计了相关权值矩阵,并采用马氏距离进行相似性度量。实验结果表明,该算法具有较高的检索率。

The content-based image retrieval on histogram is analyzed and a novel image retrieval algorithm is proposed based on bit-plane distribution entropy. Firstly, the image is divided into eight bit-planes by the image bit-plane-code. Then, an entropy vector is constructed by computing the entropy of the four significant planes which contain most of the structural information of the image. Meantime, the gray-code of bit-planes is used to avoid the effect of changes in the image intensity values on bit-planes. Finally the Mahalanobis distance is adopted to measure the similarity because of the correlation between the concerned vectors after designing the correlation-weighted matrix. Experimental results show that the proposed method has sound retrieval performance.