Scholarlink

<p>In order to explore the fitting ability of unmanned aerial vehicle (UAV) multispectral sensing to sugarcane phenotypic parameters, main sugarcane varieties GL05136 and GT42 in Guangxi were selected as the research objects, and multiple nitrogen fertilizer levels were set to carry out a field experiment. Combining the UAV multispectral data of sugarcane canopy and field data of phenotypic parameters in different growth stages from 2022 to 2023, the response characteristics of UAV multispectral sensing data to various phenotypic parameters of sugarcane in different growth stages were studied, and various modeling schemes were adopted to construct estimation models for phenotypic spectral characteristic parameters of sugarcane. The results showed that: (1) From the tillering stage to maturity stage, the SPAD, LAI and plant height of the same variety of sugarcane first increased and then decreased with different treatments, while sugar content continued to increase. The differences in phenotypic parameters among groups generally increased first and then decreased. The phenotypic parameters of ratoon cane were mostly better than newly planted cane. (2) The correlation between sugarcane phenotypes and multispectral vegetation indexes was high in the stem elongation stage and the early maturity stage, but low in other growth stages. (3) Modeling scheme S1 considering years, varieties and growth stages had the highest estimation accuracy for sugar content and SPAD, while modeling scheme S2 considering varieties and growth stages had the highest estimation accuracy for LAI and plant height. The differences in the fitting effects between S1 and S2 for SPAD, LAI and plant height were relatively small, while the difference in the fitting effect between S1 and S2 for sugar content was greater. (4) The best-fitting models of S1 and S2 for sugarcane phenotypes were mainly based on quadratic terms. The red-edge and near-infrared bands were more sensitive to changes in sugar content and SPAD, and the vegetation indexes constructed based on the red-edge and near-infrared bands achieved better estimation effects on sugar content and SPAD.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Sugarcane Phenotypic Parameters Modeling Using Unmanned Aerial Vehicle Multispectral Data

  • 期刊论文

  • 发表日期：

    2025年9月3日

查看全文
  • Yanli Chen,
  • Lu Huang,
  • Mingzhi Li,
  • Ming Sun,
  • Meihua Ding,
  • Lining Huang,
  • Ruisheng Ma,
  • Chi Liang,
  • Xin Yang,
  • Cheng Chen

摘要

In order to explore the fitting ability of unmanned aerial vehicle (UAV) multispectral sensing to sugarcane phenotypic parameters, main sugarcane varieties GL05136 and GT42 in Guangxi were selected as the research objects, and multiple nitrogen fertilizer levels were set to carry out a field experiment. Combining the UAV multispectral data of sugarcane canopy and field data of phenotypic parameters in different growth stages from 2022 to 2023, the response characteristics of UAV multispectral sensing data to various phenotypic parameters of sugarcane in different growth stages were studied, and various modeling schemes were adopted to construct estimation models for phenotypic spectral characteristic parameters of sugarcane. The results showed that: (1) From the tillering stage to maturity stage, the SPAD, LAI and plant height of the same variety of sugarcane first increased and then decreased with different treatments, while sugar content continued to increase. The differences in phenotypic parameters among groups generally increased first and then decreased. The phenotypic parameters of ratoon cane were mostly better than newly planted cane. (2) The correlation between sugarcane phenotypes and multispectral vegetation indexes was high in the stem elongation stage and the early maturity stage, but low in other growth stages. (3) Modeling scheme S1 considering years, varieties and growth stages had the highest estimation accuracy for sugar content and SPAD, while modeling scheme S2 considering varieties and growth stages had the highest estimation accuracy for LAI and plant height. The differences in the fitting effects between S1 and S2 for SPAD, LAI and plant height were relatively small, while the difference in the fitting effect between S1 and S2 for sugar content was greater. (4) The best-fitting models of S1 and S2 for sugarcane phenotypes were mainly based on quadratic terms. The red-edge and near-infrared bands were more sensitive to changes in sugar content and SPAD, and the vegetation indexes constructed based on the red-edge and near-infrared bands achieved better estimation effects on sugar content and SPAD.