Image Classification for CBIR Using Neural Network

Abstract

The recent development of computing hardware has resulted in a rapid increase of visual information such as databases of images. To successfully utilize this increasing amount of data, we need effective ways to process it. Content-based image retrieval utilizes the visual content of images directly in the process of retrieving relevant images from a database. The retrieval is based on visual features such as the colors, textures, shapes, and spatial relations the image contains rather than traditional textual keywords. This work is to cluster similar images in a large, unannotated image database. In which for each image a set of average RGB color is calculated and a SOM is trained for each feature.After that images are clustered in different classes. Results have shown that they match better with the human perception. MATLAB 7 Is the tool used to implement image processing algorithms and Microsoft Access DBMS used for managing the database of image features vector. Keywords: Image classification Content-based image retrieval, image databases, self-organizing map.