Depth-layer Architecture Reconstruction From Image Collections

An image-based modeling method is presented to generate a textured 3D model of architecture with a structure of multiple floors and depth layers. In the domain of image-based architecture modeling, it is still a challenging problem to deal with architecture in multilayered structure. We propose a statistic-based top-bottom segmentation algorithm to divide the 3D point cloud generated by structure-from-motion (SFM) method into different floors. For each floor with depth layers, we present a repetition based depth-layer decomposition algorithm to separate the front and back layers. Finally, architecture components are modeled to construct a textured 3D model. Our system has the distinct advantage of producing realistic architecture models with true depth values between front and back layers, which is demonstrated by multiple examples in the paper.