摘要： 为提高基于内容的图像检索查准率，研究基于显著闭合边缘的图像检索方法。设计显著闭合边缘提取算法，利用人类视觉注意机制获取显著图，根据原始边缘图确定显著边缘图，连接显著边缘间的断裂，得到显著闭合边缘，结合边缘梯度和角度值的直方图实现图像检索。实验结果表明，该方法具有较高的查准率。

关键词: 边缘连接, 显著边缘, 显著闭合边缘, 视觉注意机制, 显著图, 边缘端点

Abstract: In order to improve retrieval precision radio of the Content-based Image Retrieval(CBIR), this paper studies image retrieval method based on salient closed edge. It designs extraction algorithm of salient closed edge, and gets the salient map by human visual attention mechanism. It confirms the salient closed edge, and obtains salient edges by linking the ruptures between salient edges. It combines with the gradient histogram and angle value histogram of edge to realize image retrieval. Experimental results show that the proposed method has higher precision ratio.

Key words: edge linking, salient edge, salient closed edge, visual attention mechanism, salient map, edge endpoint

中图分类号: 

• TN911.73