采用位平面直方图特征向量的图像检索方法

Method of image retrieval based on bit-plane histogram features vector

首先将图像分解为8个位平面,选择前4个重要位平面,求出其灰度码表示,根据每个灰度码位平面的颜色直方图,计算均值、标准偏差、偏斜度、能量、熵;综合这些特征构成名为位平面直方图特征向量的组合特征,进行图像检索。实验中采用Tonimoto相似度量函数计算图像间的相似度。该方法计算速度快,避免了图像量化造成的误检。实验结果显示了该方法的检索性能。

Firstly,the image is divided into eight bit planes.The first four most significant bit planes of the image are considered and later each 4-bit vector is converted to its corresponding reflected gray code.Histogram features of each bit plane are computed：Mean,standard deviation,skew,energy and entropy.Then,a combined vector named bit-plane histogram features vector is constructed for image retrieval.Finally,Tonimoto measure is adopted to measure the similarity between images.The method is fast and has no improper retrieval caused by image color quantization technique.Experimental results demonstrate the performance of the method.