图像检索中融合语义和底层特征方法的研究

针对基于内容的图像检索系统检索效率不高的情况，通过在低层视觉特征上提取图像的颜色和纹理特征和空间信息，综合图像的语义特征，实现了对图像数据库的检索，最后，为了提高检索效率，把相关反馈技术引入到图像检索系统中。实验证明，该方法取得了较好的检索查全率和准确率。

Research on Image Retrieval Technique Using Semantic and Low Level Features

To improve the efficiency of image retrieval based on the content, a new method is proposed using the feature extraction of figure： region-based color and texture features are extracted from the image. We can achieve the retrieval of image database by combining the characteristic of low level and semantic infomlation. The experiments show that this method improves the precision and efficiency of image retrieval.