Multi-view stereo algorithms are able to construct highly detailed 3D models from images alone. The applications range from 3D maps to entertainment, computational photography or cultural heritage archival. This book frames the multi-view stereo problem as an image/geometry consistency optimization problem.