Combining Spectral and Texture Features for Estimating Leaf Area Index and Biomass of Maize Using Sentinel-1/2, and Landsat-8 Data

Abstract

Leaf area index (LAI) and biomass are important indicators that reflect the growth status of maize. The optical vegetation indices and the synthetic-aperture radar (SAR) backscattering coefficient are commonly used to estimate the LAI and biomass. However, previous studies have suggested that spectral features extracted from a single pixel have a poor ability to describe the canopy structure. In this paper, we propose a method for estimating LAI and biomass by combining spectral and texture features. Specifically, LAI, biomass and remote-sensing data were collected from the jointing, trumpet, flowering, and filling stages of maize. Then we formed six remote-sensing feature matrices using the spectral and texture features extracted from the remote sensing data. Principal component analysis (PCA) was used to remove noise and to reduce and integrate the multi-dimensional features. Multiple linear regression (MLR) and support vector regression (SVR) methods were used to build the estimation models. Tenfold cross-validation was adopted to verify the effectiveness of the proposed method. The experimental results show that using the texture features of both optical and SAR data improves the estimation accuracy of LAI and biomass. In particular, SAR texture features greatly improve the estimation accuracy of biomass. The estimation model constructed by combining spectral and texture features of optical and SAR data achieves the best performance (highest coefficient of determination (R 2) and lowest root mean square error (RMSE)). Specifically, we conclude that the best window sizes for extracting texture features from optical and SAR data are 3 × 3 and 7 × 7, respectively. SVR is more suitable for estimating the LAI and biomass of maize than MLR. In addition, after adding texture features, we observed a significant improvement in the accuracy of estimation of LAI and biomass for the growth stages, which have a larger variation in the extent of the canopy. Overall, this work shows the potential of combining spectral and texture features for improving the estimation accuracy of LAI and biomass in maize. INDEX TERMS Maize LAI and biomass, sentinel-1/2, spectral features, SVR, texture features.

Deep Learning-Based Estimation of Crop Biophysical Parameters Using Multi-Source and Multi-Temporal Remote Sensing Observations

“…VIs extracted from optic bands are widely used to estimate crop parameters and monitor crop conditions. However, when the crop canopy is dense, optical data tend to be saturated [7]. In addition, since optical data in cloudy conditions are not helpful, SAR sensors use microwave wavelengths that can penetrate clouds and haze [1,[8][9][10][11][12][13].…”

“…Considerable researches have been conducted to estimate various crop parameters using satellite Earth observations, including RADARSAT-2 [1,17,[19][20][21], RapidEye [5,19,20,22,23], Sentinel-1 [7,17,[24][25][26][27][28], Sentinel-2 [7,25,[29][30][31][32], Landsat-5 Thematic Mapper (TM) [33,34], Landsat-7 Enhanced Thematic Mapper Plus (ETM+) [34][35][36], Landsat-8 Operational Land Imager (OLI) [7,17,31,32,[35][36][37], Worldview-2/3 [17,27,28,38,39], and MODIS [40,41].…”

“…Zhu, et al [46] utilized unmanned aerial vehicles (UAV) data to assess the ability of four MLAs, including MLR, RF, ANN, and SVR, to estimate the above-ground biomass (AGB). Luo, et al [7] utilized MLR and SVR to estimate corn LAI and biomass. Deb, et al [37] used parametric regression models and SVR to estimate agro-pastoral AGB.…”

Integration of Radiometric Ground-Based Data and High-Resolution QuickBird Imagery with Multivariate Modeling to Estimate Maize Traits in the Nile Delta of Egypt

“…Although SRIs are simple to calculate and several indices have been effective in estimating measured traits, they are constrained by their use of just a few bands and are affected by vegetation saturation or varying degrees, as well as timeliness and regional specificity [ 38 , 39 , 40 , 41 , 78 , 79 ]. Multivariate regression models such as PLSR and MIR have previously been shown to be alternative methods to SRIs for explaining the relationships between plant-measured traits and spectral reflectance and perform equally well or better than SRIs for estimating the changes in these traits [ 42 , 45 ].…”

“…Han et al [ 78 ] found that MLR based on remote-sensing spectral data produced acceptable accuracy in predicting above-ground biomass. In addition, Luo et al [ 79 ] found that both the MLR and support vector machine (SVR) models based on spectral data showed the accuracy estimation of LAI and biomass of maize at the flowering stage since the growth and development of a maize canopy reaches a plateau at flowering, and LAI and biomass no longer grow as quickly. The overall results indicate that the proposed PLSR and MLR models based on the combination of vegetation-SRIs from two methods can improve the accuracy of plant trait estimates.…”