Relevance Feedback Methods And Fp Algorithm In Content-based Image Retrieval Applications

Images and videos are two important parts in multimedia domain. At the same time, how to find what we are interested in from huge databases has been noticed by many researchers. Under these circumstances, we discussed the technique of content-based image retrieval. Firstly, we analyzed several key levels in CBIR and summarized the description schemes of the usual features such as color, texture and shape. In the normal case, we judge the content of an image from the point of our subjective visual perception. That is to say, we understand an image' s content from its semantic feature. And it is difficult for us to find the direct relationship between the low-level features and high semantic feature. So secondly, in order to shorten the distance between the low-level features and high semantic feature, we introduce relevance feedback technique into the image retrieval process. On the basis of the users' remark, we combine the strategy of weight updating and neural network self-learning to increase the precision of image retrieval.