贝叶斯分类器在图像检索中的相关应用

The Relevant Applications of Bayesian Classifier in Image Retrieval

由于图像底层特征及其本身所包含的上层语义信息的巨大差距,使得基于内容的图像检索很难取得令人满意的结果。为了提高检索正确度,人们提取了相关反馈技术并取得了一定的成功。贝叶斯分类器是一个简单有效的分类器,将其与相关反馈技术相结合应用于图像检索,可以利用用户提供的交互信息来实现对贝叶斯分类器相关参数的更新,从而提高了系统的检索能力。

As the image itself, the underlying characteristics and semantic information contained in the huge gap between the upper, making content-based image retrieval is difficult to obtain satisfactory results. In order to improve retrieval accuracy, people extracts the relevance feedback and have achieved some success. Bayesian classifier is a simple and effective classifier, its combination with relevance feedback in image retrieval, users can take advantage of the interactive information to achieve the relevant parameters of the Bayesian classifier updates to improve the system search capabilities.