Content Based Image Retrieval System Data Mining Combined K-means, Relevance Feedback Techniques

Content Based image retrieval system (CBIR) lowlevel visual image features that is color, texture, and shape are automatically extract for image descriptions and indexing purposes. In this paper popularity of network and development of multimedia technology, the established information retrieval techniques are not working efficiently according to users command in search and retrieving images from record. In today’s world there is increased need of content based image retrieval system in number of different domains such as education, medical imaging, crime anticipation, whether forecasting, secluded sensing and association of earth property. It works on the features of images like color and texture. In this paper an enhancement to basic content based image retrieval procedure with indexing maintain by using K-means clustering data mining and relevance feedback technique. The enhanced feature helps in retrieving images from large folder initially. In this process an index is practical on database of images based on clustering technique. for the duration of this process clustering paper uses features like texture, color, shape, relevance pointer and wavelet based histogram system to find similarity among the images. Based on association value the images are separated into clusters, then the new image which is to be confirmed with database is compared with these clusters and based on its likeness corresponding images in cluster are retrieved.