Index generation for satellite image retrieval

Index generation for satellite image is an important task for retrieving satellite image in data warehouse. The goal of index generation is to find a polygon that represents the geographic image area. Using too many vertices for satellite images index can decrease retrieval speed. Therefore, the aim of this research is to develop an algorithm for generating index for satellite images using small amount of vertices and having high accuracy. The key idea of the algorithm is to decrease the number of vertices representing image shapes by mean of polygon simplification. The obtained polygon is expanded in order to make its cover the image area as much as possible. The experimental results showed that the proposed algorithm generated image index with high accuracy using small number of vertices comparing with other softwares i.e., ArcGIS 10.1, Global Mapper 15 and DIQB. In addition, the result is useful and quick for retrieving data from the spatial database.