The Application Of Computer Vision In Plant Measurement

The thesis makes a study on crops' configuration inspection technology based on computer vision and image processing technology. The study involves following works:A photograph specification is proposed for the convenience of further processing. We choose VCD disk as the standard model for crop measuring.We have analyzed the common image processing techniques in computer vision aided crop inspection. We carried out many experiments and chose clustering technique performs for image segmentation, median filter for image smoothing, Hilditch algorithm for image thinning and iterated end-point fitting algorithm for corn shape fitting.A crop configuration inspection software is developed using Microsoft Visual C++.net2003.The techniques discussed above are applied in the software.We adopted OOP techniques in the software developing process. Algorithms for Segmentation, smoothing, thinning and so on are encapsulated in separate C++ classes. This will definitely shorten the latter developing process and will improve developing efficiency.This study pioneers further research and development in applicable computer vision based crop configuration inspection system, and provides data for virtual crop modeling.