Skip to Main Content
In order to contribute on the search of solving the 3D reconstruction problem we present in this paper a workflow of 3D reconstruction from multiple view images. Our system for the time being takes a set of uncalibrated images as an input and produces a depth map. We divided the workflow in three major parts which are finding corresponding points between images, depth estimation and 3D model generation. We use the global method and dynamic programming method to produce the depth maps.