抄録
In order to effectively utilize very large image data, it is necessary to be able to efficiently retrieve the required image data from the satellite image database. For this research, we consider an image analysis system to retrieve required image data from MODIS satellite data. The MODIS data was collected as part of the research for Frontier project at Tokyo University of Information Sciences. We develop an image match search system which retrieves image data similar to a target image from the satellite database, and analyzes land surface changes in time-series for the same location. For the matching image data search algorithm, we apply genetic algorithm(GA) that is a type of probabilistic search algorithm. To increase the search speed, we constructed a cluster configuration of PCs to distribute the image search process among multiple PCs. We examine the effectiveness of distributed genetic algorithm(DGA) for the distributed MODIS data search process.
