Winter Oilseed Rape LAI Inversion via Multi-Source UAV Fusion: A Three-Dimensional Texture and Machine Learning Approach.

Improving estimation of maize leaf area index by combining of UAV-based multispectral and thermal infrared data: The potential of new texture index

PDF HTML阅读 XML下载 导出引用 引用提醒 1988-2013年基于BP神经网络的植被叶面积指数遥感定量反演 DOI: 10.5846/stxb201703290547 作者: 作者单位: 南京林业大学,南京林业大学,南京林业大学,南京林业大学,南京林业大学,南京林业大学 作者简介: 通讯作者: 中图分类号: 基金项目: 国家重点研发计划资助项目（2017YFC0505505）；国家自然科学基金项目（31200534）；江苏高校优势学科建设工程资助项目；2014年江苏省普通高校学术学位研究生科研创新计划项目 Quantitative inversion of long sequential leaf area index using remote sensing based on BP neural network from 1988 to 2013 Author: Affiliation: Key Laboratory of Soil and Water Conservation and Ecological Restoration in Jiangsu Province,Collaborative Innovation Center of Sustainable Forestry in Southern China of Jiangsu Province,Nanjing Forestry University,Nanjing,Key Laboratory of Soil and Water Conservation and Ecological Restoration in Jiangsu Province,Collaborative Innovation Center of Sustainable Forestry in Southern China of Jiangsu Province,Nanjing Forestry University,Nanjing,Key Laboratory of Soil and Water Conservation and Ecological Restoration in Jiangsu Province,Collaborative Innovation Center of Sustainable Forestry in Southern China of Jiangsu Province,Nanjing Forestry University,Nanjing,Key Laboratory of Soil and Water Conservation and Ecological Restoration in Jiangsu Province,Collaborative Innovation Center of Sustainable Forestry in Southern China of Jiangsu Province,Nanjing Forestry University,Nanjing,Key Laboratory of Soil and Water Conservation and Ecological Restoration in Jiangsu Province,Collaborative Innovation Center of Sustainable Forestry in Southern China of Jiangsu Province,Nanjing Forestry University,Nanjing,Key Laboratory of Soil and Water Conservation and Ecological Restoration in Jiangsu Province,Collaborative Innovation Center of Sustainable Forestry in Southern China of Jiangsu Province,Nanjing Forestry University,Nanjing Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:叶面积指数（Leaf Area Index，LAI）高度综合了植被水平覆盖状况和垂直结构，以及枯枝落叶层厚薄和地下生物量多少，是植被影响土壤侵蚀的主要方面。区域尺度的时间序列叶面积指数揭示了区域土壤侵蚀的演化过程。因此，及时准确地掌握区域尺度上长时间序列的植被LAI，对研究土壤侵蚀动态变化与植被的关系至关重要。选择南京市1988-2013年10期遥感影像，基于反向传播（Back Propagation，BP）神经网络构建LAI反演模型，进行了长时间序列的叶面积指数反演。结合2009和2010年LAI实测值，验证与探讨了该模型的评价精度与适应性。结果表明：（1）该模型拟合度较高，2009和2010年平均相对误差、均方根误差、相关系数分别是0.2395和0.2174，0.2962和0.2581，0.7713和0.6844，各项精度评价指标均较好；（2）统计分析去除耕地后全市LAI变化，低植被覆盖（LAI<2）面积不断增加，高植被覆盖区（LAI>3）面积先减少后增加，耕地面积不断减少，符合南京市的发展变化规律；（3）主城区LAI年际变化与其他学者得到的南京市植被盖度变化趋势一致，反演结果的时序性较高。本文提出的基于反向传播神经网络模型反演长时间序列LAI是可行的，为区域尺度土壤侵蚀定量遥感监测提供新途径。 Abstract:Leaf Area Index(LAI) can reflect the horizontal coverage, vertical structure of vegetation, the thickness of the litter layer and the amount of underground biomass, which is the main aspect of vegetation affecting soil erosion. It is very important to monitor the changes in the amount of soil erosion, for useful information to guide the planning of soil and water conservation, protect the soil and water resources and control the soil erosion. Therefore, the method by which we obtain high quality and long sequential LAI at a regional scale is very important for analyzing the relationship between the dynamic changes in soil erosion and vegetation. Previous studies showed that the neural network had an incomparable superiority in terms of complex, nonlinear data fitting and pattern recognition, and had been successfully applied to inverse the LAI in Nanjing based on the multi-spectral remote sensing data derived from the Landsat 8 Operational Land Imager(OLI), four types of vegetation indices(Normalized Difference Vegetation Index, NDVI; Ratio Vegetation Index, RVI; Soil-adjusted Vegetation Index, SAVI; Modified Soil adjusted Vegetation Index, MSAVI), and measured LAI data. The results showed that the accuracy of retrieval was good. In this paper, we used the Back Propagation(BP) neural network model to inverse the LAI in Nanjing during 1988-2013 based on the data derived from Landsat 8 OLI and Landsat 5 Thematic Mapper(TM). Based on the measured values of LAI in 2009 and 2010, the evaluation accuracy and adaptability of the model were verified and discussed. The results showed that:(1) The model had a fitting of higher degree, and average relative errors(MAPE), root mean square errors(RMSE), and correlation coefficients(R) in 2009 and 2010 of 0.2395 and 0.2174, 0.2962 and 0.2581, and 0.7713 and 0.6844, respectively. Each accuracy evaluation index was good.(2) Following statistical analysis, we found that the low vegetation coverage area(LAI < 2) exhibited an increasing trend, the high vegetation coverage area(LAI > 3) presented a first decreasing and then increasing trend, while the cultivated land area decreased with the rapid development in Nanjing.(3) To analyze accurately the LAI, we extracted the LAI in the main urban area, and found that there was a relatively high inversion value, and the inter-annual change in LAI was consistent with the change in vegetation coverage in Nanjing reported by other studies. Therefore, we could see that the BP neural network model had a high accuracy for the time series LAI inversion. It provides a new way for quantitative remote sensing monitoring of regional soil erosion. Moreover, because of other potential limiting factors, such as the errors produced by the BP neural network model, the large area of the inversion area, the complexity of vegetation types and community structure, etc., the inversion accuracy of LAI through remote sensing still needs to be explored, and the inversion method improved. We will try to establish a multi-angle LAI inversion method to construct the coupling model of LAI and soil erosion or quantitative fusion of multi-source remote sensing images in the future study. 参考文献 相似文献 引证文献

Subtropical region tea tree LAI estimation integrating vegetation indices and texture features derived from UAV multispectral images

Estimating mangrove leaf area index based on red-edge vegetation indices: A comparison among UAV, WorldView-2 and Sentinel-2 imagery

The prediction of soil moisture conditions using multispectral data from unmanned aerial vehicles (UAVs) has advantages over ground measurements in terms of costs and monitoring range. However, the prediction accuracy for moisture conditions using spectral data alone is low. In this study, relationships between water deficits and phenotypic characteristics in oats were evaluated and used to develop a UAV multispectral-based water prediction model. The vegetation indices NDRE (Normalized Difference Red Edge), CIG (Chlorophyll Index), and MCARI (Modified Chlorophyll Absorption in Reflectance Index) were highly correlated with oat yield. Based on a multipath analysis in the structural equation modeling framework, irrigation (p &lt; 0.01), leaf area index (LAI) (p &lt; 0.001), and SPAD (p &lt; 0.001) had direct positive effects on NDRE. Three distinct machine learning approaches—linear regression (LR), random forest (RF), and artificial neural network (ANN) were employed to establish predictive models between the Normalized Difference Red Edge Index (NDRE) and soil water content (SWC). The linear regression model showed moderate correlation (R2 = 0.533). Machine learning approaches demonstrated markedly superior performance (RF: R2 = 0.828; ANN: R2 = 0.810). Nonlinear machine learning algorithms (RF and ANN) significantly outperform conventional linear regression in estimating SWC from spectral vegetation indices.

Accurate leaf area index estimation in sorghum using high-resolution UAV data and machine learning models

Precise and timely leaf area index (LAI) estimation for winter wheat is crucial for precision agriculture. The emergence of high-resolution unmanned aerial vehicle (UAV) data and machine learning techniques offers a revolutionary approach for fine-scale estimation of wheat LAI at the low cost. While machine learning has proven valuable for LAI estimation, there are still model limitations and variations that impede accurate and efficient LAI inversion. This study explores the potential of classical machine learning models and deep learning model for estimating winter wheat LAI using multispectral images acquired by drones. Initially, the texture features and vegetation indices served as inputs for the partial least squares regression (PLSR) model and random forest (RF) model. Then, the ground-measured LAI data were combined to invert winter wheat LAI. In contrast, this study also employed a convolutional neural network (CNN) model that solely utilizes the cropped original image for LAI estimation. The results show that vegetation indices outperform the texture features in terms of correlation analysis with LAI and estimation accuracy. However, the highest accuracy is achieved by combining both vegetation indices and texture features to invert LAI in both conventional machine learning methods. Among the three models, the CNN approach yielded the highest LAI estimation accuracy (R 2 = 0.83), followed by the RF model (R 2 = 0.82), with the PLSR model exhibited the lowest accuracy (R 2 = 0.78). The spatial distribution and values of the estimated results for the RF and CNN models are similar, whereas the PLSR model differs significantly from the first two models. This study achieves rapid and accurate winter wheat LAI estimation using classical machine learning and deep learning methods. The findings can serve as a reference for real-time wheat growth monitoring and field management practices.

Water stress is a major factor affecting grapevine yield and quality. Standard methods for measuring water stress, such as midday stem water potential (ΨSWP), are laborious and time-consuming for intra-block variability mapping. In this study, we investigate water status variability within a 2.42-ha commercial Cabernet Sauvignon block with a standard vertical trellis system, using remote sensing (RS) tools, specifically canopy fraction-based vegetation indices (VIs) derived from multispectral unmanned aerial vehicle (UAV) imagery, as well as standard reference methods to evaluate soil and plant water status. A total of 31 target vines were monitored for ΨSWP during the whole growing season. The highest variability was at véraison when the highest atmospheric demand occurred. The ΨSWP variability present in the block was contrasted with soil water content (SWC) measurements, showing similar patterns. With spatial and temporal water stress variability confirmed for the block, the relationship between the ΨSWP measured in the field and fraction-based VIs obtained from multispectral UAV data was analysed. Four UAV flights were obtained, and five different VIs were evaluated per target vine across the vineyard. The VI correlation to ΨSWP was further evaluated by comparing VI obtained from canopy fraction (VIcanopy) versus the mean (VImean). It was found that using canopy fraction-based VIs did not significantly improve the correlation with ΨSWP (NDVIcanopyr = 0.57 and NDVImeanr = 0.53), however fractional cover (fcover) did seem to show a similar trend to plant water stress with decreasing canopy size corresponding with water stress classes. A subset of 14 target vines were further evaluated to evaluate if additional parameters (maximum temperature, relative humidity (RH), vapour pressure deficit, SWC and fractional cover) could serve as potential water stress indicators for future mapping. Results showed that the integration of NDVIcanopy and NDREmean with additional information could be used as an indicator for mapping water stress variability within a block.

Estimating the crop leaf area index (LAI) accurately is very critical in agricultural remote sensing, especially in monitoring crop growth and yield prediction. The development of unmanned aerial vehicles (UAVs) has been significant in recent years and has been extensively applied in agricultural remote sensing (RS). The vegetation index (VI), which reflects spectral information, is a commonly used RS method for estimating LAI. Texture features can reflect the differences in the canopy structure of rice at different growth stages. In this research, a method was developed to improve the accuracy of rice LAI estimation during the whole growing season by combining texture information based on wavelet transform and spectral information derived from the VI. During the whole growth period, we obtained UAV images of two study areas using a 12-band Mini-MCA system and performed corresponding ground measurements. Several VI values were calculated, and the texture analysis was carried out. New indices were constructed by mathematically combining the wavelet texture and spectral information. Compared with the corresponding VIs, the new indices reduced the saturation effect and were less sensitive to the emergence of panicles. The determination coefficient (R2) increased for most VIs used in this study throughout the whole growth period. The results indicated that the estimation accuracy of LAI by combining spectral information and texture information was higher than that of VIs. The method proposed in this study used the spectral and wavelet texture features extracted from UAV images to establish a model of the whole growth period of rice, which was easy to operate and had great potential for large-scale auxiliary rice breeding and field management research.

Canopy water interception is a key parameter to study the hydrological cycle, water utilization efficiency, and energy balance in terrestrial ecosystems. Especially in sprinkler-irrigated farmlands, the canopy interception further influences field energy distribution and microclimate, then plant transpiration and photosynthesis, and finally crop yield and water productivity. To reduce the field damage and increase measurement accuracy under traditional canopy water interception measurement, UAVs equipped with multispectral cameras were used to extract in situ crop canopy information. Based on the correlation coefficient (r), vegetative indices that are sensitive to canopy interception were screened out and then used to develop canopy interception models using linear regression (LR), random forest (RF), and back propagation neural network (BPNN) methods, and lastly these models were evaluated by root mean square error (RMSE) and mean relative error (MRE). Results show the canopy water interception is first closely related to relative normalized difference vegetation index (R△NDVI) with r of 0.76. The first seven indices with r from high to low are R△NDVI, reflectance values of the blue band (Blue), reflectance values of the near-infrared band (Nir), three-band gradient difference vegetation index (TGDVI), difference vegetation index (DVI), normalized difference red edge index (NDRE), and soil-adjusted vegetation index (SAVI) were chosen to develop canopy interception models. All the developed linear regression models based on three indices (R△NDVI, Blue, and NDRE), the RF model, and the BPNN model performed well in canopy water interception estimation (r: 0.53–0.76, RMSE: 0.18–0.27 mm, MRE: 21–27%) when the interception is less than 1.4 mm. The three methods underestimate the canopy interception by 18–32% when interception is higher than 1.4 mm, which could be due to the saturation of NDVI when leaf area index is higher than 4.0. Because linear regression is easy to perform, then the linear regression method with NDVI is recommended for canopy interception estimation of sprinkler-irrigated winter wheat. The proposed linear regression method and the R△NDVI index can further be used to estimate the canopy water interception of other plants as well as forest canopy.

Unmanned aerial vehicle (UAV)-based remote sensing (RS) possesses the significant advantage of being able to efficiently collect images for precision agricultural applications. Although numerous methods have been proposed to monitor crop nitrogen (N) status in recent decades, just how to utilize an appropriate modeling algorithm to estimate crop leaf N content (LNC) remains poorly understood, especially based on UAV multispectral imagery. A comparative assessment of different modeling algorithms (i.e., simple and non-parametric modeling algorithms alongside the physical model retrieval method) for winter wheat LNC estimation is presented in this study. Experiments were conducted over two consecutive years and involved different winter wheat varieties, N rates, and planting densities. A five-band multispectral camera (i.e., 490 nm, 550 nm, 671 nm, 700 nm, and 800 nm) was mounted on a UAV to acquire canopy images across five critical growth stages. The results of this study showed that the best-performing vegetation index (VI) was the modified renormalized difference VI (RDVI), which had a determination coefficient (R2) of 0.73 and a root mean square error (RMSE) of 0.38. This method was also characterized by a high processing speed (0.03 s) for model calibration and validation. Among the 13 non-parametric modeling algorithms evaluated here, the random forest (RF) approach performed best, characterized by R2 and RMSE values of 0.79 and 0.33, respectively. This method also had the advantage of full optical spectrum utilization and enabled flexible, non-linear fitting with a fast processing speed (2.3 s). Compared to the other two methods assessed here, the use of a look up table (LUT)-based radiative transfer model (RTM) remained challenging with regard to LNC estimation because of low prediction accuracy (i.e., an R2 value of 0.62 and an RMSE value of 0.46) and slow processing speed. The RF approach is a fast and accurate technique for N estimation based on UAV multispectral imagery.

Above-ground biomass (AGB) and the leaf area index (LAI) are important indicators for the assessment of crop growth, and are therefore important for agricultural management. Although improvements have been made in the monitoring of crop growth parameters using ground- and satellite-based sensors, the application of these technologies is limited by imaging difficulties, complex data processing, and low spatial resolution. Therefore, this study evaluated the use of hyperspectral indices, red-edge parameters, and their combination to estimate and map the distributions of AGB and LAI for various growth stages of winter wheat. A hyperspectral sensor mounted on an unmanned aerial vehicle was used to obtain vegetation indices and red-edge parameters, and stepwise regression (SWR) and partial least squares regression (PLSR) methods were used to accurately estimate the AGB and LAI based on these vegetation indices, red-edge parameters, and their combination. The results show that: (i) most of the studied vegetation indices and red-edge parameters are significantly highly correlated with AGB and LAI; (ii) overall, the correlations between vegetation indices and AGB and LAI, respectively, are stronger than those between red-edge parameters and AGB and LAI, respectively; (iii) Compared with the estimations using only vegetation indices or red-edge parameters, the estimation of AGB and LAI using a combination of vegetation indices and red-edge parameters is more accurate; and (iv) The estimations of AGB and LAI obtained using the PLSR method are superior to those obtained using the SWR method. Therefore, combining vegetation indices with red-edge parameters and using the PLSR method can improve the estimation of AGB and LAI.

In the cultivation of maize, the leaf area index (LAI) serves as an important metric to determine the development of the plant. Unmanned aerial vehicles (UAVs) that capture RGB images, along with random forest regression (RFR), can be used to indirectly measure LAI through vegetative indices. Research using these techniques is at an early stage, especially in the context of maize for silage. Therefore, this study aimed to evaluate which vegetative indices have the strongest correlations with maize LAI and to compare two regression methods. RFR, ridge regression (RR), support vector machine (SVM), and multiple linear regression (MLR) regressions were performed in Python for comparison using images obtained in an area cultivated with maize for silage. The results showed that the RGB spectral indices showed saturation when the LAI reached 3 m2 m−2, with the VEG (vegetable index), COM (combination), ExGR (red–green excess), and TGI (triangular greenness index) indices selected for modeling. In terms of regression, RFR showed superior performance with an R2 value of 0.981 and a root mean square error (RMSE) of 0.138 m2 m−2. Therefore, it can be concluded that RFR using RGB indices is a good way to indirectly obtain the LAI.

Improving estimation of LAI dynamic by fusion of morphological and vegetation indices based on UAV imagery

Synergistic monitoring of cotton verticillium wilt using GF-2 satellite and UAV data: A dual-branch feature fusion approach