Hierarchical Histograms: A New Representation Scheme for Image based Data Retrieval

This paper proposes a new image representation scheme for database indexing. The application domain is "image-based" indexing as opposed to "content-based" indexing. In other words, given a database of images and associated data, we want to search the database for a known image so as to retrieve the data associated with the image. A concrete example is to search a database of aerial images so as to retrieve the GPS coordinates of a known scene. Our representation is based on a hierarchy of color histograms and offers a compromise between methods based on a global histogram and methods based on multiple historgrams corresponding to segmented regions. Our method retains positional information better than the above schemes and naturally leads to a multi-level comparison strategy where mismatches are quickly discarded at higher levels. Our method offers a simple, easily implementable solution to the specific problem of image-based indexing.