中文摘要：

提出了一种基于图划分和图像搜索引擎的图像标注改善算法，通过对待标注图像的候选标注词进行去噪处理，提高标注的准确性．算法的核心思想是将候选标注词作为图的顶点，将标注词间的相关度作为边的权值，从而把图像标注改善问题转换为图划分问题．用2个参数对标注词间的相似度进行加权处理后计算出边的权值：参数1是根据图像搜索引擎返回结果计算出的候选标注词与待标注图像视觉特征之间的相关度；参数2是候选标注词在待标注图像所属页面中的重要程度，此参数仅适用于Web图像．然后，用启发式最大割算法对构造出的图进行二划分，最后从划分出的2个顶点集中选择其一作为最终标注．实验结果表明，对比已有方法，使用本算法对非Web图像和Web图像进行标注改善后，最终的标注结果都更加准确．

英文摘要：

Automatic image annotation has been an active research direction due to its great importance in content-based image retrieval（CBIR）. However, the results of existing image annotation methods are still far from practical. Therefore, it is of vital importance to design a high-performance apf）roach which could refine the initial annotations. This paper presents a novel algorithm to solve image annotation refinement problem （IAR） by graph partition and image search engine. Our algorithm focuses on pruning the noisy words in candidate annotation set to enhance image annotation performance. The main idea of the proposed algorithm lies in that candidate annotations are served as graph vertices, and the relevance between two candidate annotations is used to construct the edge weight. Then, the image annotation refinement problem can be converted to the weighted graph partition problem. The edge weight is the annotation similarity weighted by two parameters. Parameter 1 is the relationship between candidate annotation and image visual features, and parameter 2 refers to the importance of candidate annotation in host Web page. Next, we compute max cut of the graph using a heuristic algorithm. After the graph is bi-partitioned, one of the two vertex sets is chosen as final annotations. Experimental results on non-Web images and Web images show that our algorithm outperforms the existing image annotation refinement techniques.