Abstract: The CBIR tends to index and retrieve images based on their visual content. CBIR avoid many evils connected with conservative ways of retrieve images by keywords. Thus, a rising interest in the area of CBIR has been known in current years. The arrangement of a CBIR system mainly depends on the particular image illustration and similarity matching function working. The CBIR tends to index and retrieve images based on their visual content. CBIR avoids many problems associated with traditional ways of retrieving images by keywords. Thus, a growing interest in the area of CBIR has been established in recent years. The performance of a CBIR system mainly depends on the particular image representation and similarity matching function employed. So a new CBIR system is proposed which will provide accurate results as compare to previous developed systems. Soft system will be used in this system. Based Image recovery system which evaluates the similarity of each image in its data accumulate to a query image in terms of various visual features and return the image with desired range of similarity. To develop and put into practice an efficient feature extraction NN and SVM to extract features according to data set using Auto calculate the feature weight by neural network. The precision and recall graph in gui according to the retrieved contents of the images from the datasets. To Apply back propagation or feed forward algorithm for neural network classification. To calculate cross relationship and apply weakening model for feature matching.
Keywords: CBIR, KNN, ABIR, precision, Recall etc.