基于多反馈累积综合的新闻图像检索方法

A News Image Retrieval Method on the Basis of Multi-Feedbacks Cumulative and Synthesizing Method

通过分析新闻图像检索的应用特点,提出了一种多反馈、累积的图像检索方法.在贝叶斯分类模型的基础上,构造图像特征的分类方法,进一步得到图像的分类概率;设计多个反馈指标上的概率综合公式及先验概率的累积修正方法.实验结果表明,所提出的算法是有效的,并具有较好的性能.

According to application features of news image retrieval, a multi-feedbacks cumulative and synthesizing retrieval method is proposed. On the basis of Bayesian theory, the classifying method about image features is constructed. Then, the classifying probability of image can be gained. Through constructing probability synthesizing formula about multi-feedbacks and cumulative updating formula about prior probability, a relevance feedback algorithm based on Bayesian theory is proposed. Experimental results show that the proposed algorithm is effective and has higher performance.