Image-Based Tree Modeling

In this paper, the authors propose an approach for generating 3D models of natural-looking trees from images that has the additional benefit of requiring little user intervention. While the approach is primarily image-based, they do not model each leaf directly from images due to the large leaf count, small image footprint, and widespread occlusions. Instead, they populate the tree with leaf replicas from segmented source images to reconstruct the overall tree shape. In addition, they use the shape patterns of visible branches to predict those of obscured branches. They demonstrate the approach on a variety of trees.