Estimating the Growth Indices and Nitrogen Status Based on Color Digital Image Analysis During Early Growth Period of Winter Wheat

Ben Zhao; Yonghui Zhang; Aiwang Duan; Zhandong Liu; Junfu Xiao; Zugui Liu; Anzhen Qin; Dongfeng Ning; Sen Li; Syed Tahir Ata-Ul-Karim

doi:10.3389/fpls.2021.619522

Estimating the Growth Indices and Nitrogen Status Based on Color Digital Image Analysis During Early Growth Period of Winter Wheat

Front Plant Sci. 2021 Apr 8:12:619522. doi: 10.3389/fpls.2021.619522. eCollection 2021.

Authors

Ben Zhao¹, Yonghui Zhang², Aiwang Duan¹, Zhandong Liu¹, Junfu Xiao¹, Zugui Liu¹, Anzhen Qin¹, Dongfeng Ning¹, Sen Li¹, Syed Tahir Ata-Ul-Karim^{3

4}

Affiliations

¹ Key Laboratory of Crop Water Use and Regulation, Ministry of Agriculture, Farmland Irrigation Research Institute, Chinese Academy of Agricultural Sciences, Xinxiang, China.
² School of Computer Engineering, Weifang University, Weifang, China.
³ Institute for Sustainable Agro-Ecosystem Services, The University of Tokyo, Tokyo, Japan.
⁴ Department of Global Agricultural Sciences, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Tokyo, Japan.

Abstract

The non-destructive estimation of plant nitrogen (N) status is imperative for timely and in-season crop N management. The objectives of this study were to use canopy cover (CC) to establish the empirical relations between plant growth indices [shoot dry matter (SDM), leaf area index (LAI), shoot N accumulation (SNA), shoot nitrogen concentration (SNC)], and CC as well as to test the feasibility of using CC to assess N nutrition index (NNI) from Feekes 3 to Feekes 6 stages of winter wheat. Four multi-locational (2 sites), multi-cultivars (four cultivars), and multi-N rates (0-300 kg N ha^-1) field experiments were carried out during 2016 to 2018 seasons. The digital images of the canopy were captured by a digital camera from Feekes 3 to Feekes 6 stages of winter wheat, while SDM, LAI, SNA, and SNC were measured by destructive plant sampling. CC was calculated from digital images developed by self-programmed software. CC showed significant correlations with growth indices (SDM, LAI, and SNA) across the different cultivars and N treatments, except for SNC. However, the stability of these empirical models was affected by cultivar characteristics and N application rates. Plant N status of winter wheat was assessed using CC through two methods (direct and indirect methods). The direct and indirect methods failed to develop a unified linear regression to estimate NNI owing to the high dispersion of winter wheat SNC during its early growth stages. The relationships of CC with SDM, SNC and NNI developed at individual growth stages of winter wheat using both methods were highly significant. The relationships developed at individual growth stages did not need to consider the effect of N dilution process, yet their stability is influenced by cultivar characteristics. This study revealed that CC has larger limitation to be used as a proxy to manage the crop growth and N nutrition during the early growth period of winter wheat despite it is an easily measured index.

Keywords: canopy cover; computer vision technology; nitrogen diagnosis; nitrogen nutrition index; winter wheat.