Based On The Design Of A Web Image Search Engine And Database Systems

With the Popularization and development of Internet,the number of image data grows extremely fast,and how to retrieve image effieiently and quikcly beeomes an important issue in the field of image's application. In the early period,images were often labelled as a series of keywords manually , and then retrieve the keywords of the image instead of retrieving the image directly.However, there are some significant shortcomings with this approach.On the one hand,the workload is too huge,and on the other hand,the keywords labelled of the image are subjective.So the CBIR (Content-based image retrieval) has emerged to be one of the hot research areas in image domain.CBIR is a kind of technique for image retrieval on the basis of automatically extracting visual features such as color, texture, and shape etc.CBIR extracts visual features as the retrieval features,such as color,texture,shape ete.Generally,a sample image is needed when a user uses the CBIR system,which can extract the features of the sample image,then compare with the features of other images in the data base and show the result to the user.As an important vision information of image, the calculation of color characteristic and texture is steady, this has been broadly used in the Content Based Image Retrieval .Although many feature- extracting methods and similarity measurements have been raised,they are not mature enough to get good retrieval results.New technology of CBIR researching is required to improve retrieval performance.This dissertation briefly summarizes CBIR system,and researches some key techniques of the image retrieval which specially focuses on feature extraetion.In this paper we proposes a new method for image retrieval,it makes use of both content feature and semantic feature of images.So It can solve the problem that the content feature of image dosen't agree with it's semantic feature.The experimental result of special image database including images classified in 20 kinds demonstrates its retrieval precision is increased over than the image retrieval method based on content feature alone.The experimental result shows the limitation of the method CBIR. Based on the first method proposed in this dissertation,a new method for image retrieval based on dimensionality reduction. The new method with dimensionality reduction does not only reduces the space taken up by the image features ,but also cuts down the redundance information of image features.If the image database is pre-classified before image search,the efficiency and Accuracy can further improve.The experimental result of the same images database demonstrates its retrieval Precision is inereased comparing with the first method be proposed in this dissertation,and it can increase the retrieval Precision remarkably,comparing with global HSV colorhistogram image retrieval method.In the end,a prototype image retrieval system is constructed,which realizes the approaches proposed before.The experiment With the a prototype image retrieval system and image database classified in 20 kinds shows that the image retrieval algorithms proposed in this paper are efficient and effective.