Optimization of Lai Estimation Method Based on Smartphones with Fisheye Lens

Author:

Zhu Lichen1,Guan Peng1,Liu Weiping1,Zheng Yili1

Affiliation:

1. School of Technology, Beijing Forestry University, Beijing 100083, P.R. China

Abstract

Leaf area index (LAI) is an important biological factor reflecting vegetation growth and forest ecosystem. LAI can be used to obtain plant health status, carbon cycle, and surrounding ecological environment effectively. In this study, the smartphone was equipped with a fisheye lens, and the optimization method was used to estimate LAI, which was compared with digital hemispherical photography (DHP) to investigate the possibility of the new method for LAI estimation. The hemispherical image was divided into blocks, and the optimized Otsu method was used for algorithm segmentation, which can effectively distinguish vegetation from the sky. Concurrently, when the gap fraction inversion LAI was performed, the linear inversion algorithm was improved based on single-angle inversion, and the LAI was obtained by inversion through the linear fitting of the mul-tiangle gap fraction. The experimental sample was located in Olympic National Forest Park in Beijing. Three coniferous mixed forests and three broadleaved forests were selected from the experimental sample. LAI measurements from smartphones were compared with those from DHP. In the samples for mixed coniferous forests, the values for coefficients of determination R^2 were 0.835, 0.802, and 0.809, and root mean square errors (REMS) were 0.137, 0.120, and 0.147. For the broadleaf forest samples, the values for R² were 0.629, 0.679, and 0.758, and REMS were 0.144, 0.135, and 0.137. The R^2 and RMES for the overall data was 0.810 and 0.134, respectively, and a good agreement between the LAI measurements from the proposed method and those from the DHP supports an accurate estimation. The results show that the use of a fisheye lens on a smartphone can effectively and accurately obtain tree canopy LAI. This provides a fast and effective new method to measure LAI of forest vegetation near the ground, which is of great significance for studying the interaction between plant growth status, ecological environment, and phenological changes.

Publisher

North Atlantic University Union (NAUN)

Subject

Electrical and Electronic Engineering,Signal Processing

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3