Functional-structural plant modeling and plant phenotyping require the measurement of geometric features in specimens. This data acquisition, plant digitizing, is usually performed manually, in invasive or even destructive ways, or using expensive laser scanning equipment. Computer vision based 3D reconstruction is an accurate and low cost alternative for the digitizing of plants not presenting a dense canopy. State of the art works in structure from motion and multiple view stereo are able to produce accurate 3D models for specimens presenting sparse canopies. Three-dimensional triangular meshes can be computed from a set of non-calibrated images, modeling plants and accurately representing their leaves and branches. Point cloud processing can be employed to segment and classify structures in the model, allowing the estimation of, for example, (i) parametric representations for shapes, (ii) surface orientation (leaf normal) and (iii) leaf area.