Deformed Lattice Detection in Real-World Images Using a Novel Mean-Shift Belief

The software library is based on our paper, "Deformed Lattice Detection in Real-World Images Using Mean-Shift Belief Propagation", TPAMI special issue OCTOBER 2009 (Vol. 31, No. 10). It detects, automatically, various repetitive patterns as deformed 2D lattices in real photos with cluttered background. Examples include windows on buildings, fruits on display, draped cloth, or see-through wiry structures. The lattice detection problem is formulated as a spatial, multi-target tracking problem, solved within an MRF framework using a novel mean-shift belief propagation method. The model is evaluated and compared with state of the art existing methods on a wide range of images that contain near-regular textures (NRT).