采用形态学边界特征的医学图像检索

Medical Image Retrieval Method Using Boundary Shape Feature

特征提取是基于内容的图像检索(CBIR)中的关键步骤,如何有效提取反映高层语义的图像特征对于医学图像的检索至关重要.提出一种基于边界形状特征的医学图像检索方法.该方法首先通过多尺度形态学方法检测图像边界点,然后对边界图像进行形状特征提取,构建边界的形状密度直方图,最后通过相似性匹配实现医学图像检索.实验结果证明了所提取的边界形状特征在医学图像检索中的有效性,通过对比实验给出了结果分析和进一步的研究思路.

Feature extraction is one of the most crucial steps in content-based image retrieve （ CBIR ）. How to extract image features which reflect the high level semantics of an image is very import.ant for medical image retrieval. In accordance with this issue, there proposed a method based on boundary shape feature for medical image retrieval in this paper. Firstly image edges were detected by a multi-scale morphological gradient algorithm. Then shape feature was extracted from the obtained boundary image and shape-density histogram was constructed. Lastly medical image retrieval was executed according on similar computing using the extracted shape fea- tures. The results of experimentation showed that the proposed algorithm has been applied to medical image retrieval with promising effect. Moreover, the result analysis and the future work were given.