基于多重分割捆绑特征的目标图像检索

针对基于目标的图像检索（OBIR）领域中,传统的视觉关键词方法忽略了局部特征之间的空间关系信息,导致检索准确度不高的问题,提出一种基于多重分割捆绑特征的目标图像检索方法.通过对图像进行多重分割,各分割区块用它所包含的尺度不变特征变换(SIFT)特征集合来描述,生成包含空间关系信息的捆绑特征；根据视觉关键词词库匹配捆绑特征,并提出一种改进的相似性度量方法计算捆绑特征相似度,再将该相似度作为权重融入到视觉关键词方法的向量空间模型中,计算图像相似度并进行排序.结果表明,该方法能够有效利用局部特征之间的空间关系信息,在保证检索效率的同时,显著提高检索准确度.

Bundling features with multiple segmentations for object-based image retrieval

In the area of object based image retrieval, the traditional visual words (VW) based methods neglected the spatial relationship among local features, resulting in the low accuracy. To overcome this problem, a novel method of bundling features with multiple segmentations was proposed. In our method, images were multiple segmented, and all segmentations were described by SIFT features fallen inside the area to generate bundling features. The bundling features were matched based on the VW vocabulary. An improved similarity metric was presented to measure the similarity between matched bundling features, and the degree of similarity was infused into the vector space model of VW method, to calculate the image similarity. Experiments show that the proposed method can exploit the space relationships among local features, and improve the retrieval accuracy greatly with no significant reduction in the efficiency.