Interactive lattice visualization of huge image database by formal concept analysis

An interactive visualization system based on formal concept analysis (FCA) is proposed, in order to extract interesting rules with respect to large scale image database. The proposed system employs HSB color space that is suitable for human sensitivity to draw the lattice structure of images, and as the attribute of FCA, quantization of the HSB color space is used. Furthermore, an interactive functionality of the proposed system can increase the readability of the visualized lattice and it may be better for uses to parse the essence of data structure. The proposed system is developed by using JAVA based program language Processing on the computer (CPU = 2.13GHz, MM = 2GB), and a visualization experiment is performed. Through the experiment using huge number of images selected from ubiquitous home database produced by National Institute of Information and Communication Technology (NICT), it is confirmed the effectiveness of the proposed system, to grasp the structure in perspective and to discover new relationship.