Image mosaic or image montage is an image made up of many other images. In this paper, we propose a methodology of generating image mosaics. Our image mosaic generating system divides an input image into many tiles; and then for each tile, it fetches the image with the most similar content from an image database and replaces the tile with the image. CBIR techniques and the approach which we propose are used in unison to select the image with the most similar content to a tile. Due to the large size of our image collection, a compact visual feature representation, as well as its distance metric, has been used to reduce the storage and processing time. There is a set of parameters used in our system such as the number of tiles and the size of the image database/tiles which controls the procedure of image mosaic generating.