本项目立足于满足现代农业对高性能作物三维测量技术的迫切需求，针对条纹投影法对作物三维测量时存在的被测表面光响应与光分布规律复杂、视场零重合点云拼接等问题，从作物多光谱响应机理、条纹图编码、拼接约束构建等环节入手开展创新性研究：探索彩色叶片的多光谱响应特性，构建作物色彩-投影光源谱段-相位精度关联模型，揭示簇生叶片条纹投影测量时串扰影响规律；提出自适应恒定相位技术，制定背景项、幅值优化的条纹图自适应编码测量方案；基于叶片自然纹理特征，联合点云质量分布，构造拼接评价函数，提出一种逐点加权拼接算法。本项目研究有助于合理选用投影光源谱段，抑制串扰效应，解决条纹图优化编码难题，突破视场零重合点云高精度拼接瓶颈，提高条纹投影法对作物的全视角三维测量性能，促进其在农业领域的推广应用，推动我国农业测量技术的提升。

This project aims at meeting the urgent demand on high-performance three-dimensional (3D) measurement techniques in modern agriculture. Considering the problems that the laws of the light response and the light spatial distribution are very complex, and the 3D point clouds to be registered might be isolated to each other when a crop is measured with fringe projection profilometry (FPP), the multi-spectrum light response mechanism, the coding strategy of fringe pattern, and the construction of registration constraint are researched in the project. Firstly, the multi-spectrum response properties of color leaves are researched, the correlation model between the color, the spectrum, and the phase accuracy is constructed, and the phase disturbance law caused by the adjacent leaves is revealed; secondly, the adaptive uniform-phase technique is put forward, and then the strategy to optimize the fringe background and fringe modulation is presented; lastly, a pixel by pixel weighted registration algorithm is proposed based on properties of the texture of a leaf and the quality distribution of the 3D point clouds. These researches help to select the suitable spectrum for the light source of a projection unit, suppress the adverse impact of the disturbance, solve the problem in optimally coding the fringe pattern, break the bottleneck in registering point clouds without mutual field of view, enhance the measurement performance of FPP to the crop, facilitate its application in agriculture, and eventually boost the technological level in the agricultural measurement of our country.