A Fast Visual Modeling Plant Based on Images

This paper proposes a fast visual modeling plant based on images,reconstruct an approximate plant model.Firstly,based on knowledge of growth patterns for plant,using image segmentation and 3D reconstruction to construct the plant skeleton,modeling leaves,and growing the remainder of the plant.Then the system produce the realistic plant model images based on image synthesis and validation.It is not necessary to acquire the complex production rules of L-systems.The proposed method provides a highly control over final shape,resulting in realistic reconstruction.