Method for three-dimensional reconstruction of plant population morphological structure

Abstract

The invention discloses a method for three-dimensional reconstruction of a plant population morphological structure, relating to the technical field of three-dimensional graph generation. The method comprises the following steps of: S1, measuring canopy morphological character parameter data of a plant population to be reconstructed; S2, constructing a plant population canopy morphological character statistical model according to the morphological character parameter data; S3, acquiring position information and internode number information of each plant of the plant population to be reconstructed; S4, reconstructing a three-dimensional model of each plant in the plant population to be reconstructed according to the internode number information acquired in the step S3 and the canopy morphological character statistical model constructed in the step S2; and S5, according to the position information, moving the three-dimensional model of each plant to a corresponding position of a three-dimensional scene to finish the three-dimensional reconstruction of the plant population to be reconstructed. By the method, the three-dimensional reconstruction of the plant population can be accurately and simply performed.

Description

technical field [0001] The invention relates to the technical field of three-dimensional graphics generation, in particular to a method for three-dimensional reconstruction of plant population morphological structure. Background technique [0002] In order to establish the three-dimensional morphological structure of plant populations, some researchers have proposed a simulation method for plant morphological structure based on process models. Typical simulation models for the process of plant morphological structure changes include L-system, GreenLab, etc. The above methods mainly establish a mathematical model describing the process of morphological structure changes in the process of plant growth through observation and experimental data analysis, including the relationship between morphological structure changes and environmental factors. On this basis, a set of environmental condition parameters can be input to simulate the change process and results of plant morphology...

AI Technical Summary

Problems solved by technology

The main problems of this method are: (1) Since there are many branches and leaves in the plant canopy, twigs and leaves are often difficult to reconstruct directly from the scan data, so this method is generated by applying some random rules or knowledge of plant morphology twigs and leaves, so as to achieve visually "real" reconstruction, but the three-dimensional plant canopy structure reconstructed in this way is still quite different from real plants, especially the density of leaves, spatial orientation, leaf area, etc. In fact, there are large errors, which are difficult to apply to the research and analysis of canopy light distribution characteristics and plant type characteristics; (2) The amount of plant 3D point cloud data obtained by 3D laser scanners is generally very large, and often a plant The point cloud data has reached millions of scales. It is a huge challenge to identify each organ of the plant and restore the original shape and structure of the plant from these large amounts of unordered point cloud data. There are many, and the occlusion problem is serious, resulting in the loss of some organ data in the 3D point cloud acquired by the scanner

For a single organ, this method is easier to obtain the shape of the organ, but for the whole plant, there are still many problems to be solved to automatically restore the geometric shape of the plant from the image sequence of the plant

For example, the segmentation and automatic recognition of organs in images, the restoration of occluded organs, etc. are complex problems, and there is no satisfactory solution yet.

That is to say, this method is difficult to solve the problem of restoring the occluded organs in the image. Since there are many organs in the canopy of the plant population, mutual occlusion is very common, and the accuracy of the 3D model obtained by the image-based reconstruction method is difficult to guarantee.

Images