In this thesis, a new methodology of tomato plants structural model (SM) construction is presented. The methodology is based on the concepts from 3D reconstruction modeling methods, integrated with modern laser scanners and CAD tools to result in an architecturally accurate and human manipulable tomato plant model reflecting crop growth and management over time. A laser scanner is used for plant measurement and the resulting data are intelligently simplified. Critical data is extracted and exported to a CAD system as reference points to form a parametric model construction. A 2D corner detection algorithm is developed and applied for reference point extraction in 3D scans. A hierarchical structure is used to decompose the tomato plants into weekly-growth-sections (WGS) to relate age with plant growth geometry and solving occlusion and complexity problems. A mathematical plant growth model is developed to define geometry and topology of the plant components, WGS and the plant as a whole.;Key words. Tomato Plant Modelling, CAD, Parametric model, Point cloud, Rule-based plant model, 3D reconstruction, Laser scanning, corner detection.
展开▼
