Accuracy and efficiency are two major issues in designing Content-Based Image Retrieval (CBIR) systems. In this paper an indexing structure for retrieval of remote sensing images on the basis of shape and texture is developed. The objects are extracted using differential Morphological Segmentation (DMP) algorithm. The objects are then encoded into bitmaps and represented in a grid descriptor which is very intuitive. The similar objects are then grouped using DBSCAN clustering algorithm which is used to cluster multiscale data sets. Finally an indexing structure called EBB tree is developed to efficiently retrieve objects based on shape and texture.