基于三维点云的植物叶片重建及其面积估算  

Plant Leaf Reconstruction and Its Area Estimation Based on 3D Point Cloud

作　　者：陈柱 杨君[1,2] CHEN Zhu;YANG Jun(Engineering Research Center of Metallurgical Automation and Measurement Technology of the Ministry of Education,Wuhan University of Science and Technology,Wuhan Hubei 430081,China;School of Information Science and Engineering,Wuhan University of Science and Technology,Wuhan Hubei 430081,China)

机构地区：[1]武汉科技大学冶金自动化检测技术教育部工程研究中心,湖北武汉430081 [2]武汉科技大学信息科学与工程学院,湖北武汉430081

出　　处：《传感技术学报》2022年第3期349-354,共6页Chinese Journal of Sensors and Actuators

基　　金：国家自然科学基金项目(61801338)。

摘　　要：为提高植物叶片面积测量的准确度,提出了一种植物叶片三维重建补偿方法。该方法首先使用多角度拍摄植物叶片的方法来获取图像;其次,通过运动恢复结构(SFM)算法、聚类多视角立体(CMVS)算法和基于面片的多视角立体(PMVS)算法处理图像并生成三维点云;然后,对点云进行去噪、分割、填补、三角网格化处理;最后,对叶片面积进行估测。实验结果表明,本文方法测量叶片面积的准确度与扫描法接近,并且能解决由于叶片重叠产生的叶面积测量不准确的问题。To improve the accuracy of plant leaf area measurement,a compensation method of plant leaf three-dimensional reconstruction is proposed.Firstly,the images are captured by method of photographing plant leaves from multiple angles.Secondly,the images are processed and three-dimensional point clouds are generated by structure from motion(SFM)algorithm,the clustering multi-view stereo(CMVS)algorithm and the patch-based multi-view stereo(PMVS)algorithm.Then,the point clouds are denoised,segmented,filled and triangulated.Finally,the leaf areas are estimated.Our experimental results show that our method can accurately measure blade area which is close to the scanning method,and it can solve the problem of inaccurate measurement of leaf area caused by blade overlap.

关 键 词：三维重建 点云填补 运动恢复结构 稀疏点云 稠密点云 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]