Two fast illumination invariant image retrieval methods for scenes comprising textured objects with variable illumination are introduced. Both methods are based on texture gradient modelled by efficient set of random field models. We developed the illumination insensitive measures for textured images representation and compared them favorably with steerable pyramid and Gabor features in the illumination invariant BTF texture recognition. Problem formulation Illumination invariant texture retrieval using single training image per class with unknown illumination direction. Textures from the University of Bonn BTF database (81 different illuminations for a fixed view per texture). Proposed Solution ? Multiscale texture gradient based on MRF type of BTF texture representations: