智慧农业（中英文）2025，Vol.7Issue(6)：149-160,12.DOI:10.12133/j.smartag.SA202508026
基于物理约束PROSAIL-cGAN的冬小麦LAI光谱样本增强与反演方法
Physics-Constrained PROSAIL-cGAN Approach for Spectral Sample Augmentation and LAI Inversion of Winter Wheat
摘要
Abstract
[Objective]The leaf area index(LAI)is a key biophysical parameter that reflects the canopy structure and photosynthetic ca-pacity of crops.However,the inversion of winter wheat LAI from remote sensing data is often constrained by the limited availability of field measurements,leading to insufficient model generalization.Although radiative transfer model(RTM)-based simulations can expand the sample size,discrepancies between simulated and observed spectra persist due to simplified canopy and soil parameteriza-tions.Conversely,purely data-driven generative models such as generative adversarial networks(GANs)can enhance sample diversity but often produce physically inconsistent samples in the absence of biophysical constraints.To address these issues,a physics-con-strained PROSAIL-cGAN(conditional generative adversarial network)spectral sample augmentation method was proposed that inte-grated the PROSAIL model with cGAN to improve the accuracy and robustness of LAI inversion under small-sample conditions,gen-erate physically realistic spectral-parameter pairs and provide reliable data support for remote sensing-based monitoring of winter wheat growth.[Methods]The study area was located in Zouping city,Shandong Province,a major winter wheat production region within the Huang-Huai-Hai Plain.A total of 133 field samples were collected during the jointing stage in April 2025 using an LAI-2200C canopy analyzer,with synchronous canopy spectra acquired.A Sentinel-2A Level-2A image from April 15,2025,served as the remote sensing source,comprising 13 bands resampled to a spatial resolution of 10 m.The dataset was divided into training(70%)and validation(30%)subsets,with LAI values ranging from 1.646 to 7.505.The proposed method combined the PROSAIL radiative transfer model with a conditional GAN framework.First,PROSAIL was employed to simulate canopy reflectance and corresponding biophysical parameters,including chlorophyll content(Cab),carotenoid content(Car),brown pigment content(Cbrown),equivalent water thickness(Cw),dry matter content(Cm),LAI,and leaf inclination distribution(LIDFa).A multi-layer perceptron(MLP)surrogate mod-el was then trained to approximate the forward mapping of PROSAIL,enabling differentiability for integration with deep learning ar-chitectures.The cGAN generator received random noise and physical parameters as conditional inputs to produce corresponding cano-py reflectance,while the discriminator jointly evaluated authenticity and physical consistency.During adversarial training,physical constraints were incorporated into the generator's loss function to ensure biophysical realism.The generated samples were subsequent-ly filtered based on parameter ranges and discriminator confidence scores.Kernel density overlap between real and generated LAI dis-tributions was used to quantify their statistical consistency.Finally,the enhanced dataset was used to train random forest(RF)and ex-treme gradient boosting(XGBoost)regression models for the LAI inversion.Model performance was assessed using the coefficient of determination(R2),root mean square error(RMSE),and mean absolute error(MAE),and compared with three baselines:1)field-mea-sured modeling,2)the PROSAIL lookup table(LUT)method,and 3)cGAN-only augmentation.[Results and Discussions]The surro-gate MLP model accurately reproduced PROSAIL-simulated spectra,achieving R2 0.817,RMSE 0.008 5,and MAE 0.005 5,confirm-ing its feasibility as a differentiable physical proxy.The cGAN-based augmentation achieved a LAI distribution overlap of 0.806 with the measured samples,whereas the PROSAIL-cGAN improved the overlap to 0.827,demonstrating enhanced physical realism and sample diversity.Model comparisons revealed substantial differences in performance.The LUT-based inversion yielded only R2 0.353 0 and RMSE 1.284 0,reflecting its limited adaptability to spectral heterogeneity.Direct regression using field data improved ac-curacy(R2=0.680 1 for XGBoost and 0.648 8 for RF).Incorporating cGAN-generated samples further enhanced model accuracy(R2 0.745 0 for RF and 0.739 0 for XGBoost).The PROSAIL-cGAN-enhanced RF model achieved the best overall performance,with R2 0.848 8,RMSE 0.540 9,and MAE 0.293 7.The sample-size sensitivity analysis demonstrated that as the number of field samples in-creased from 27 to 106,R2 improved from 0.546 2 to 0.848 8 and RMSE decreased from 1.024 3 to 0.540 9.When the sample size ex-ceeded 79,model performance stabilized,indicating strong robustness.Spatial mapping results showed that LAI values were higher in the central and northern regions(4～7)and lower in the southern mountainous areas(1.5～4),consistent with variations in soil fertility and field management practices.These findings validate the model's applicability for regional-scale monitoring of crop growth.[Con-clusions]This study developed a physics-constrained PROSAIL-cGAN spectral sample augmentation method for winter wheat LAI in-version.By integrating a radiative transfer model,a conditional generative network,and a differentiable surrogate,the method effec-tively generated physically consistent and diverse spectral-parameter samples under small-sample conditions.The PROSAIL-cGAN-based RF model achieved a relatively high inversion accuracy,outperforming traditional LUT and field-only approaches.The pro-posed method successfully mitigated small-sample limitations,ensured physical interpretability,and improved model generalization.It provides a robust framework for the remote sensing inversion of crop canopy parameters,supporting precision agriculture and dy-namic monitoring of crop growth.Future work will focus on optimizing sample generation strategies,integrating multi-temporal satel-lite data and additional physiological parameters,and coupling with deep or semi-supervised learning techniques to further enhance scalability and applicability across crops and regions.关键词
物理约束/PROSAIL/条件生成对抗网络/样本增强/叶面积指数/遥感反演Key words
physical constraint/PROSAIL/conditional generative adversarial network/sample augmentation/leaf area index/remote sensing inversion分类
农业科技引用本文复制引用
LU Yihang,DONG Wen,ZHANG Xin,YAN Ruoyi,ZHANG Yujia,TANG Tao..基于物理约束PROSAIL-cGAN的冬小麦LAI光谱样本增强与反演方法[J].智慧农业（中英文）,2025,7(6):149-160,12.基金项目
国家重点研发计划项目(2021YFB3901300) National Key Research and Development Program of China(2021YFB3901300) （2021YFB3901300）
