基于分数维的彩色图像检索研究

为使图像检索结果接近人类感知,提高单一特征的检索精度,提出了一种综合特征的图像检索新方法。通过计算图像在HSV颜色空间中各个通道的分形维数,并将分形维数非线性量化,使得分形维数与粗糙度呈近似线性比例关系,以便于与图像的颜色直方图特征相结合实现图像相似匹配计算。实验表明,该方法能够达到较好的检索结果,检索效果在一定程度上符合人眼感知。

Research of color image retrieval based on fractal dimension

In order to improve the precision of image retrieval,and make the retrieval result being similar to the human visual sense,a new retrieval method was proposed.Through calculating the fractal dimension feature and the color feature,which are both extracted in the HSV color space,the fractal dimension in nonlinear way is quantified,to make sure that the relation of fractal dimension and the roughness to be approximately linearity.The experiments show that the proposed method can get an effective search results,it is some of similar to the human visual sense.