This method extracts 3D information from a large number of images, taken from various positions. On each image, specific spots like edges or ridges have to be highlighted and tagged manually by mouse clicking. The shifts in space, of all highlighted structures between the different pictures, brings about the 3D shape of the edited object.
The method is not reliable, error prone and only works well for simple, large objects. It is not recommendable for plants, with their complex structures and lack of contrast.