Leaf area index estimation of a row-planted eggplant canopy using wide-angle time-lapse photography divided according to view-zenith-angle contours
Leaf area index estimation of a row-planted eggplant canopy using wide-angle time-lapse photography divided according to view-zenith-angle contours
复制标题
DOI：
10.1016/j.agrformet.2022.108930
复制
发表时间：
2022-05
影响因子：
6.2
通讯作者：
K. Nomura;M. Saito;M. Kitayama;Yudai Goto;Koyo Nagao;Hiromi Yamasaki;T. Iwao;T. Yamazaki;I. Tada;M. Kitano
中科院分区：
文献类型：
--
作者：
K. Nomura;M. Saito;M. Kitayama;Yudai Goto;Koyo Nagao;Hiromi Yamasaki;T. Iwao;T. Yamazaki;I. Tada;M. Kitano
在全年园艺水果生产中，叶面积指数 (LAI;L) 是适当作物管理的有用指标。在这项研究中，使用广角延时数码相机来估计行栽茄子冠层在整个生长期间的L。顶篷上方放置了六台广角延时摄影机；三台相机指向最低点，另外三台相机指向与最低点成57.5°倾斜角度的不同方位角方向（即沿行、与行成40°角、垂直于行）。L是基于间隙分数理论使用三种方法估计的：（1）裁剪法，其中仅使用每张照片的一部分来计算L， (2)全图像法，即使用每张照片的整体来计算L；(3)轮廓法，即根据视角天顶角轮廓将每张照片划分为区域，并通过区域L的加权平均来估计L。使用裁剪方法估计的L值根据相机位置和方向存在严重偏差，而使用其他两种方法获得的值受相机位置和方向的影响较小，表明宽视角的优势。与全图像方法相比，轮廓方法提高了估计精度。在球形叶角分布的假设下，最低点摄影倾向于高估L，而57.5°角摄影与异速生长获得的参考L值更一致。通过应用具有优化参数的β叶角分布函数，使用最低点摄影获得的估计的准确性得到了显着提高。这项研究强调了（1）广角摄影用于估计中耕作物冠层 L 的适用性，（2）轮廓方法的优点，以及（3）使用适当的叶角分布的重要性。
In year-round horticultural fruit production, the leaf area index (LAI;L) is a useful indicator for proper crop management. In this study, wide-angle time-lapse digital cameras were applied to estimate theLof a row-planted eggplant canopy throughout an entire growth period. Six wide-angle time-lapse cameras were placed above the canopy; three cameras pointed to the nadir, and the other three pointed at 57.5°-tilted angles from the nadir with different azimuthal orientations (i.e., along the row, at a 40° angle from the row, and perpendicular to the row).Lwas estimated based on the gap-fraction theory using three methods: (1) the crop method, in which only a part of each photograph was used to calculateL, (2) the whole-image method, in which the entirety of each photograph was used to calculateL, and (3) the contour method, in which each photograph was divided into zones according to the view-zenith-angle contours andLwas estimated through weighted averaging of the zonalLs. TheLvalues estimated using the crop method were heavily biased depending on the camera position and orientation, whereas the values obtained with the other two methods were less affected by the camera position and orientation, indicating the advantage of a wide viewing angle. Compared to the whole-image method, the contour method improved theLestimation accuracy. Under the assumption of a spherical leaf-angle distribution, nadir-looking photography tended to overestimateL, whereas 57.5°-angled photography was more consistent with the allometrically obtained referenceLvalue. The accuracy of theLestimations obtained using nadir-looking photography was improved considerably by applying the beta leaf-angle distribution function with optimized parameters. This study highlights the (1) applicability of wide-angle photography for estimating theLof a row-crop canopy, (2) advantage of the contour method, and (3) importance of using a proper leaf-angle distribution.