Evaluation of an IGA-based image retrieval system using wavelet coefficients

We evaluate the performance of an interactive genetic algorithm (IGA)-based image retrieval system with a subjective test. First, the IGA-based system that retrieves images based on wavelet analysis is introduced. Second, a psychological scale space is constructed to quantitatively express mental images to handle subjective retrieval. Third, the IGA-based system is compared with a random search-based system using the psychological space and a subjective test. The two sign tests and a statistical test have shown that the IGA-based system is significantly quicker in image retrieval than a random search-based image retrieval system.