Abstract:Soybean lodging identification on breeding purpose relies on manual measurements and visual inspection by the breeders, which is inefficient, time-consuming, and very subjective to human errors. From the past few years, remote sensing image has become one of the most popular tools for crop lodging identification due to the tools provides features for building machine learning models for the lodging prediction of the cultivars. The goal of this study was to investigate the potential use of UAV-based imagery in identifying lodging of soybean breeding lines that can be used to select lodging tolerance soybean genotypes using UAV-based imagery and machine learning methods. An UAV platform equipped with an RGB (red-green-blue) camera was used to collect the imagery data at flight height of 30m from 1266 four-row plots of soybean breeding fields at the reproductive stage. Soybean lodging scores were visually assessed and measured by the breeders and the scores were grouped into four classes namely, non-lodged, moderate lodged, high lodged and severe lodged. Fourteen textural features such as angular second moment, contrast, correlation, variance, inverse difference moment, sum average, sum variance, sum entropy, entropy, difference variance, difference entropy, information measure of correlation I, information measure of correlation II and maximal correlation coefficient were extracted to classify the soybean lodging scores that are associated with the image features. Using Random Forest (RF) Recursive Feature Elimination (RFE) method was used to select important features. In the preprocessing step, Synthetic Minority Oversampling Technique (SMOTE) and Edited Nearest Neighbors (ENN) method were employed to treat the imbalanced lodging dataset. Later, random forest (RF), K-nearest neighbor (KNN), artificial neural network (ANN) models were developed to classify the four classes of soybean lodging using the image derived texture features. Among all the classification models developed, ANN classification using normalized image feature showed higher precision, recall and overall accuracy (95%) to classify the soybean classes. Whereas RF also showed promising overall accuracy of 94% with minimum misclassification rate of 3%, 3%, 3% and 1%, for non-lodging, medium lodging, high lodging, and severe lodging classes, respectively. The higher classification accuracy of soybean lodging classes reflected and demonstrate that the use of UAV based imagery techniques has the high potentiality to identify and select the non-lodging soybean genotypes.

Monitoring of maize lodging using multi-temporal Sentinel-1 SAR data

Monitoring of rice lodging using Sentinel-1 data

A Novel Approach to Estimate Maize Lodging Area with PolSAR Data

Classification of Maize Lodging Types Using UAV-SAR Remote Sensing Data and Machine Learning Methods

Detection and Analysis of Degree of Maize Lodging Using UAV-RGB Image Multi-Feature Factors and Various Classification Methods

Assessing Maize Lodging Severity Using Multitemporal UAV-based Digital Images

An improved approach to estimating crop lodging percentage with Sentinel-2 imagery using machine learning

A Decision-Tree Approach to Identifying Paddy Rice Lodging with Multiple Pieces of Polarization Information Derived from Sentinel-1

Quantitative Identification of Maize Lodging-Causing Feature Factors Using Unmanned Aerial Vehicle Images and a Nomogram Computation

Evaluating how lodging affects maize yield estimation based on UAV observations

In-season Prediction of Maize Stem Lodging Risk Using an Active Canopy Sensor

Estimation of canopy nitrogen nutrient status in lodging maize using unmanned aerial vehicles hyperspectral data

Coupling GEDI LiDAR and Optical Satellite for Revealing Large‐Scale Maize Lodging in Northeast China

Identifying Corn Lodging in the Mature Period Using Chinese GF-1 PMS Images.

Identification of Early Lodging Resistance of Maize by Hyperspectral Imaging Technology

Precise identification of maize in the North China Plain based on Sentinel-1A SAR time series data

Evaluating the Patterns of Maize Development in the Hetao Irrigation Region Using the Sentinel-1 GRD SAR Bipolar Descriptor

Quantification of Lodging Scores of Soybean Breeding Lines Using UAV-based Imagery and Machine Learning

An Explainable XGBoost Model Improved by SMOTE-ENN Technique for Maize Lodging Detection Based on Multi-Source Unmanned Aerial Vehicle Images

Estimation of soybean yield parameters under lodging conditions using RGB information from unmanned aerial vehicles

35. In-season Prediction of Maize Lodging Characteristics Using an Active Crop Sensor