植物冠层的形态结构与植物群体的生理功能和产量均存在直接的关系，体现了植物群体对肥水、光照和温度等调控措施或环境条件变化的响应。由于叶片的角度分布与叶片含水量和冠层光截获直接相关，是反应作物生理状态的重要指标，也是表述冠层形态结构的关键参数，因此植物冠层叶片角度的连续在线监测也是近年来国内外农业信息化领域的热点问题。本项目针对双目立体视觉技术在农作物生长监测上易受环境和作物本身影响等问题，以黄瓜为对象，重点研究自然光照条件下植物叶片图像分割方法，以及植物叶片的立体匹配与三维重建方法，并进而实现冠层叶片倾角和方位角的自动计算。从而为评价不同作物品种的冠层特征、对比不同生产管理措施对作物群体的影响等应用提供技术手段，促进农学、信息学、机电工程学之间的学科交叉和融合。

The morphological structure of plant canopies is close link with both its physiological function and yield. The morphological structure is also the result of plant populations adapting to environmental conditions, such as water, light and temperature. The leaf angle distribution in plan canopies is directly related to leaf water content and light interception, makes it an important indicator for reflecting crop physiological state, and a key parameter for descripting canopy morphological structure. As such in recent years the continuous online monitoring of leaf angle in plant canopies became a hot issue in agricultural information. The project selects binocular stereo vision technology as the quantitative acquirement method of leaf angle in plant canopies, against the problem of binocular stereo vision technology on crops growth monitoring is vulnerable to environment and crop itself. The major concern of this project is leaf image segmentation under Illumination invariant conditions, stereo matching and 3D Reconstruction of leaves under complex context, with the aim of establishing a quantitative model for measuring blade angle in plant canopies under natural field environment. The project wish to provide a fundamental approach and operative technology for evaluating high-yielding varieties, high-yield cultivation system.