基于组合特征的图像检索方法仿真研究

摘要

In order to improve the accuracy of image retrieval, this paper put forward an image retrieval method based on combined features. Firstly, the color features of the images, the texture characteristics and shape characteristics were extracted, then the principal component analysis was used to eliminate the redundant information of the image characteristics vector. At last, we used Euclidean distance to measure the similarity among the images, so as to realize the image retrieval. The image database SIMPLIcity was used for the simulation to test the performance of this method. The simulation results show that, compared with the traditional identification methods, this method can overcome the deficiency of traditional methods, improve image retrieval accuracy, and speed up the image retrieval speed.%研究图像检索优化问题,由于采用元数据检索,是对图像进行点与点匹配,传统单一特征或简单多特征组合方法难以正确描述图像,导致图像检索准确率较低.为了提高图像检索准确率,提出一种组合特征的图像检索方法.首先分别提取图像的颜色特征、纹理特征和形状特征,然后采用主成分分析消除各图像特征向量的冗余信息,最后采用欧式距离方式对图像之间的相似度进行度量,从而实现图像检索.采用图像数据库SIMPLIcity对方法性能进行仿真测试,仿真结果表明,相对于传统识别方法,组合特征方法很好地克服了传统方法不足,不仅提高了图像检索准确率,同时加快图像检索速度,是一种有效的图像检索识别方法.