Accurate crop-type classification using multi-temporal optical and multi-polarization SAR data in an object-based image analysis framework

Accurate crop-type classification is a challenging task due, primarily, to the high within-class spectral variations of individual crops during the growing season (phenological development) and, second, to the high between-class spectral similarity of crop types. Utilizing within-season multi-temporal optical and multi-polarization synthetic aperture radar (SAR) data, this study introduces a combined object- and pixel-based image classification methodology for accurate crop-type classification. Particularly, the study investigates the improvement of crop-type classification by using the least number of multi-temporal RapidEye (RE) images and multi-polarization Radarsat-2 (RS-2) data utilized in an object- and pixel-based image analysis framework. The method was tested on a study area in Manitoba, Canada, using three different classifiers including the standard Maximum Likelihood (ML), Decision Tree (DT), and Random Forest (RF) classifiers. Using only two RE images of July and August, the proposed method results in overall accuracies (OAs) of about 95%, 78%, and 93% for the ML, DT, and RF classifiers, respectively. Moreover, the use of only two quad-pol images of RS-2 of June and September resulted in OAs of 92%, 75%, and 90% for the ML, DT, and RF classifiers, respectively. The best classification results were achieved by the synergistic use of two RE and two RS-2 images. In this case, the overall classification accuracies were 97% for both ML and RF classifiers. In addition, the average producer’s accuracies of 95% and 96% were achieved by the ML and RF classifiers, respectively, whereas the average user accuracy was 94% for both classifiers. The results indicated promising potentials for rapid and cost-effective local-scale crop-type classification using a limited number of high-resolution optical and multi-polarization SAR images. Very accurate classification results can be considered as a replacement for sampling the agricultural fields at the local scale. The result of this very accurate classification at discrete locations (approximately 25 × 25 km frames) can be applied in a separate procedure to increase the accuracy of crop area estimation at the regional to provincial scale by linking these local very accurate spatially discrete results to national wall-to-wall continuous crop classification maps.     

Analysis of multi-frequency and multi-polarization SAR data for wetland mapping in Hamoun-e-Hirmand wetland

ABSTRACT

The complex, dynamic and narrow boundaries between vegetation types make wetland mapping challenging. Hereafter the case study of the Hamoun-e-Hirmand wetland is considered by analysing eight Synthetic Aperture Radar (SAR) Images acquired in dry and wet periods with three wavelengths (X-band ~ 3 cm, C-band ~ 6 cm, and L-band ~ 25 cm), three polarizations (HH, VV and VH), and four incidence angles (22°, 30°, 34° and 53°). Then, the Support Vector Machine (SVM) classification method was applied to classify TerraSAR-X, Sentinel-1, and ALOS-PALSAR images. The final wetland land cover map was created by combining the classification results obtained from each sensor. In the case in question, results show that TerraSAR-X (X-band, HH-53°) and Sentinel-1 data (C-band, VV-34°) were useful for determining the flooded vegetation area in the wet period. This is crucial for the conservation of water bird habitats since flooded vegetation is an ideal environment for the nesting and feeding of water birds. PALSAR data (L-band in both HH and VH polarizations, 30°) were capable of separating the classes of vegetation density in the wetland. In the dry period, Sentinel-1 (VV and VH, 34°) and TerraSAR-X (HH, 22° and 53°) had higher potential in land cover mapping than PALSAR (HH and VH, 30°). Based on these results, Sentinel-1 in VV and VH provides the highest ability to discriminate between dry and green plants. TerraSAR-X is better for separating meadow and bare land. The results obtained in this paper can reduce the ambiguity in selecting satellite data for wetland studies. The results can also be used to produce more accurate data from satellite images and to facilitate wetland investigation, conservation and restoration.     

Potential of SAR sensors TerraSAR-X, ASAR/ENVISAT and PALSAR/ALOS for monitoring sugarcane crops on Reunion Island

Multi-temporal TerraSAR-X, ASAR/ENVISAT and PALSAR SAR data acquired at various incidence angles and polarizations were analyzed to study the potential of these new spaceborne SAR systems for monitoring sugarcane crops. The sensitivity of different radar parameters (wavelength, incidence angles, and polarization) to sugarcane growth stages was analyzed to determine the most suitable radar configuration for better characterisation of sugarcane fields and in particular the monitoring of sugarcane harvest.Correlation between backscattered signals and crop height was also carried out. Radar signal increased quickly with sugarcane height until a threshold height, which depended on radar wavelength and incidence angle. Beyond this threshold, the signal increased only slightly, remained constant, or even decreased. The threshold height is higher with longer wavelengths (L-band in comparison with C- and X-bands) and higher incidence angles (~ 40° in comparison with ~ 20°).The radar backscattering coefficients (σ°) were also compared to the Normalized Difference Vegetation Index (NDVI) calculated from SPOT-4/5 images. Results showed a high correlation between the behaviors of σ° and NDVI as a function of sugarcane crop parameters. A decrease in NDVI for fully mature sugarcane fields due to drying of the sugarcane (water stress) was also observed in the radar signal. This decrease in radar signal was of the same order as the decrease in radar signal after the sugarcane harvest. In general, it is more suitable to monitor the sugarcane harvest using high incidence angles regardless of the radar wavelength. SAR data in L- and C-bands showed an ambiguity between the signals of ploughed fields and those of fields in vegetation because of the high sensitivity of the radar signal at these wavelengths to surface roughness of bare soils. Indeed, sometimes the radar signal of ploughed fields was of the same order as that of harvested or mature sugarcane fields. Results showed better discrimination between ploughed fields and sugarcane fields in vegetation (sugarcane canopy) when using TerraSAR-X data (X-band).Concerning the influence of radar polarization, results showed that the co-polarizations channels (HH and VV) were well correlated, but had slightly less potential than cross-polarization channels (HV and VH) for the detection of the sugarcane harvest. Finally, SAR data at high spatial resolution were shown to be useful and necessary for better analysis of SAR images when the fields were of small size.     

A unified framework for crop classification in southern China using fully polarimetric,dual polarimetric,and compact polarimetric SAR data

With the development of synthetic aperture radar (SAR) techniques, various imaging modes that involve single polarimetry, dual polarimetry, full polarimetry (FP), and compact polarimetry (CP) have been proposed and applied to SAR systems. This article attempts to introduce a unified framework for crop classification in southern China using FP, coherent HH/VV, and CP data. By analysing the polarimetric response from different land-cover types (including rice, banana trees, sugarcane, eucalyptus, water, and built-up areas in the experimental site) and by exploring the similarities between data in these three modes, a knowledge-based characteristic space is created and a unified classification framework is presented. Time-series data acquired by TerraSAR-X over the Leizhou Peninsula, southern China, are used in our experiments. The overall classification accuracies for data in the FP and coherent HH/VV modes are approximately 95%, and for data in the CP mode, the accuracy is 91%, which suggest that the proposed classification scheme is effective. Compared with the Wishart Maximum Likelihood (ML) classifier, the proposed method provides approximately 5.64%, 7.30%, and 6.48% higher classification accuracies in the FP, HH/VV, and circular transmit and dual circular receive modes, respectively.     

Sentinel-1-based flood mapping: a fully automated processing chain

This article presents an automated Sentinel-1-based processing chain designed for flood detection and monitoring in near-real-time (NRT). Since no user intervention is required at any stage of the flood mapping procedure, the processing chain allows deriving time-critical disaster information in less than 45 min after a new data set is available on the Sentinel Data Hub of the European Space Agency (ESA). Due to the systematic acquisition strategy and high repetition rate of Sentinel-1, the processing chain can be set up as a web-based service that regularly informs users about the current flood conditions in a given area of interest. The thematic accuracy of the thematic processor has been assessed for two test sites of a flood situation at the border between Greece and Turkey with encouraging overall accuracies between 94.0% and 96.1% and Cohen’s kappa coefficients (κ) ranging from 0.879 to 0.910. The accuracy assessment, which was performed separately for the standard polarizations (VV/VH) of the interferometric wide swath (IW) mode of Sentinel-1, further indicates that under calm wind conditions, slightly higher thematic accuracies can be achieved by using VV instead of VH polarization data.     

基于面向对象的多源卫星遥感影像玉米倒伏面积提取

风灾引起的玉米倒伏可能导致玉米大量减产,利用遥感技术准确监测玉米倒伏面积与空间分布信息对灾情的评估非常重要。利用Planet和Sentinel-2影像分别结合面向对象与基于像元方法提取研究区玉米倒伏,同时评估了不同影像特征（光谱特征、植被指数和纹理特征）与不同分类方法（支持向量机法SVM、随机森林法RF和最大似然法MLC）对玉米倒伏提取精度的影响。结果表明：(1)使用高空间分辨率的Planet影像进行玉米倒伏提取的精度普遍高于Sentinel-2影像;(2)从分类精度和面积精度来看,Planet影像的光谱特征+植被指数+均值特征结合面向对象RF分类,总体精度和Kappa系数分别为93.77%和0.87,面积的平均误差最低为4.76%;(3)采用Planet和Sentinel-2影像结合面向对象分类提取玉米倒伏精度高于基于像元分类。研究不仅分析了面向对象方法的优势,还评估了使用不用影像数据结合面向对象方法的适用性,可以为遥感提取作物倒伏相关研究提供一定的借鉴。     

Mapping rice areas with Sentinel-1 time series and superpixel segmentation

Rice is the single most important crop for food security in Asia. Knowledge about the distribution of rice fields is also relevant in the context of greenhouse-relevant methane emissions, disease transmission, and water resource management. Copernicus Sentinel-1 provides the first openly available archive of C-band SAR (synthetic aperture radar) data at high spatial and temporal resolution. We developed one of the first methods that shows the potential of this data for accurate and timely mapping of rice-growing areas. We used superpixel segmentation to create spatially averaged backscatter time series, which is robust to speckle and reduces the amount of data to process. This method has been applied to six study sites in different rice-growing regions of the world and achieved an average overall accuracy of 0.83.     

Mapping of rice-cropping pattern and cultural type using remote-sensing and ancillary data: a case study for South and Southeast Asian countries

A geospatial database on the spatial distribution of rice areas and rice cultural types of major rice-producing countries of South and Southeast Asia has been developed in this study using remote-sensing and ancillary data sets. Multitemporal SPOT VGT normalized difference vegetation index (NDVI) data for the period 2009–2010 were used for the analysis. The classification was performed adopting ISODATA clustering to build a non-agricultural area mask followed by rice area mapping. The derived rice area was stratified by logical modelling of ancillary data sets into five rice cultural types: irrigated wet, upland, flood-prone, drought-prone, and deep-water. The uniqueness of this study is a synergistic approach based solely on single-source, high-temporal remote-sensing data coupled with ancillary data, which demonstrate the application of SPOT VGT NDVI data in building a geospatial database for rice crops over a wide spatial extent. This approach was adopted for cost effectivity as the study extent was vast and thus lacking ground truth information. Comparison of the derived rice area against the reported literature values for validation yielded a good correlation (linear coefficient of determination, R² = 0.95–0.99). The high-temporal resolution NDVI data enabled effective characterization of vegetation phenology. The derived spatial outputs can be used in various studies associated with the assessment of greenhouse gas emissions from paddy fields, change detection, and inputs to crop simulation models, which are significantly related to different rice cultural types.     

A Review of: “Exploration of the Solar System by Infrared Remote Sensing”. By R. A. HANEL,B. J. CONRATH,D. E. JENNINGS and R. E. SAMUELSON. (Cambridge: Cambridge University Press, 1992.) [Pp. 458.] Price £75.00.

Crop classification is a key issue for agricultural monitoring using remote-sensing techniques. Synthetic aperture radar (SAR) data are attractive for crop classification because of their all-weather, all-day imaging capability. The objective of this study is to investigate the capability of SAR data for crop classification in the North China Plain. Multi-temporal Envisat advanced synthetic aperture radar (ASAR) and TerraSAR data were acquired. A support vector machine (SVM) classifier was selected for the classification using different combinations of these SAR data and texture features. The results indicated that multi-configuration SAR data achieved satisfactory classification accuracy (best overall accuracy of 91.83%) in the North China Plain. ASAR performed slightly better than TerraSAR data acquired in the same time span for crop classification, while the combination of two frequencies of SAR data (C- and X-band) was better than the multi-temporal C-band data. Two temporal ASAR data acquired in late jointing and flowering periods achieved sufficient classification accuracy, and adding data to the early jointing period had little effect on improving classification accuracy. In addition, texture features of SAR data were also useful for improving classification accuracy. SAR data have considerable potential for agricultural monitoring and can become a suitable complementary data source to optical data.     

Integration of multitemporal/polarization C‐band SAR data sets for land‐cover classification

This paper investigates the potential of multitemporal/polarization C‐band SAR data for land‐cover classification. Multitemporal Radarsat‐1 data with HH polarization and ENVISAT ASAR data with VV polarization acquired in the Yedang plain, Korea are used for the classification of typical five land‐cover classes in an agricultural area. The presented methodologies consist of two analytical stages: one for feature extraction and the other for classification based on the combination of features. Both a traditional SAR signal property analysis‐based approach and principal‐component analysis (PCA) are applied in the feature extraction stage. Special concerns are in the interpretation of each principal component by using principal‐component loading. The tau model applied as a decision‐level fusion methodology can provide a formal framework in which the posteriori probabilities derived from different sensor data can be combined. From the case study results, the combination of PCA‐based features showed improved classification accuracy for both Radarsat‐1 and ENVISAT ASAR data, as compared with the traditional SAR signal property analysis‐based approach. The integration of PCA‐based features based on multiple polarization (i.e. HH from Radarsat‐1, and both VV and VH from ENVISAT ASAR) and different incidence angles contributed to a significant improvement of discrimination capability for dry fields which could not be properly classified by using only Radarsat‐1 or ENVISAT ASAR data, and thus showed the best classification accuracy. The results of this case study indicate that the use of multiple polarization SAR data with a proper feature extraction stage would improve classification accuracy in multitemporal SAR data classification, although further consideration should be given to the polarization and incidence angle dependency of complex land‐cover classes through more experiments.     

基于CASI影像的黑河中游种植结构精细分类研究

基于CASI高光谱影像资料,计算出NDVI和纹理数据并综合进行SVM(Support Vector Machine)分类,3种信息的组合形成4种分类方案,是为了探讨CASI数据在种植结构精细分类中的应用潜力,为定量研究和监测提供数据基础。数据在分类前利用同步ASD数据和CE\|318数据进行了辐射定标和大气校正。分类结果与地面实际调查数据对比验证结果表明:① 4种分类结果均与地面实际调查情况基本一致,并分别取得了96.78%、97.21%、88.00%、88.38% 的分类精度和0.9676、0.9691、0.8674、0.8716的Kappa系数;② CASI数据信息丰富,在植被的精细分类方面具有很大的应用潜力;③ 结合空间特征信息和NDVI数据可以有效地提高分类精度。     

基于SAR纹理信息的农作物识别研究——以农安县为例

以吉林省农安县为研究区,以Sentinel-1B双极化数据为数据源,提取出典型农作物玉米、大豆、水稻的多个纹理特征值,筛选出最佳农作物识别纹理信息参数,结合eCognition软件中的规则库,充分挖掘SAR数据中农作物纹理特征包含的属性信息,构建决策树,基于面向对象分类方法对典型农作物进行提取,通过SAR农作物提取结果...     

基于Sentinel-2的闪电河流域农作物分类研究

以内蒙古闪电河流域为研究区,基于Sentinel2光学遥感影像结合随机森林和支持向量机算法,采用3种方案:基于像元的分类方法、面向对象的分类方法及改进的基于像元分类与面向对象分割相结合的集成方法,对研究区内的农作物进行精细提取.结果表明:①基于随机森林采用基于像元的方法进行分类,所有地类的总体精度为97.8％,Kapp...     

Comparison of support vector machine,artificial neural network,and spectral angle mapper algorithms for crop classification using LISS IV data

The Resourcesat-2 is a highly suitable satellite for crop classification studies with its improved features and capabilities. Data from one of its sensors, the linear imaging and self-scanning (LISS IV), which has a spatial resolution of 5.8 m, was used to compare the relative accuracies achieved by support vector machine (SVM), artificial neural network (ANN), and spectral angle mapper (SAM) algorithms for the classification of various crops and non-crop covering a part of Varanasi district, Uttar Pradesh, India. The separability analysis was performed using a transformed divergence (TD) method between categories to assess the quality of training samples. The outcome of the present study indicates better performance of SVM and ANN algorithms in comparison to SAM for the classification using LISS IV sensor data. The overall accuracies obtained by SVM and ANN were 93.45% and 92.32%, respectively, whereas the lower accuracy of 74.99% was achieved using the SAM algorithm through error matrix analysis. Results derived from SVM, ANN, and SAM classification algorithms were validated with the ground truth information acquired by the field visit on the same day of satellite data acquisition.     

Methodology to classify rice cultural types based on water regimes using multi-temporal RADARSAT-1 data

This study presents a methodology to classify rice cultural types based on water regimes using multi-temporal synthetic aperture radar (SAR) data. The methodology was developed based on the theoretical understanding of radar scattering mechanisms with rice crop canopy, considering crop phenology and variation in water depth in the rice field, emphasizing the sensitivity of SAR to crop geometry and water. The logic used was the characteristic decrease in SAR backscatter that is associated with the puddled or transplanted field due to specular reflection for little exposure of crop, with increase in backscatter as the crop growth progresses due to volume scattering. Besides, the multiple interactions between SAR and vegetation/water also lead to an increase in backscatter as the crop growth progresses. Classification thresholds were established based on the information provided by each pixel in each image, the pixel's typical temporal behaviour due to crop phenology and changing water depth in rice field and their corresponding SAR signature. Based on this logic, the study site (i.e. South 24 Paraganas district, West Bengal) was classified into three major rice cultural types, namely shallow water rice (SWR; 5 cm ≤ water depth ≤ 30 cm), intermediate water rice (IWR; 30 cm ≤ water depth ≤ 50 cm) and deep water rice (DWR; water depth > 50 cm) during the kharif season. These three types represent most of the traditional rice-growing areas of India. The methodology was validated with the field data collected synchronously with the satellite passes. Classification results showed an overall accuracy of 98.5% (95.5% kappa coefficient) compared with a maximum-likelihood classifier (MLC) with an overall accuracy of 95.5% (84.2% of kappa coefficient) with 95% confidence interval. The relationship between field parameters, especially exposed plant height and water depth with SAR backscatter, was explored to design empirical models for each of the three rice classes. Significant relationships were observed in all the rice classes (coefficient of determination, R ², value more than 0.85) even though they had similar growth profiles but varied with water depth. The two main conclusions drawn from this study are (i) the importance of multi-temporal SAR data for the classification of rice culture types based on water regimes and (ii) the advantages and flexibility of the knowledge-based classifier for classification of RADARSAT-1 data. However, being empirical, the approach needs modification according to the current rainfall pattern and rice-growing practice.     

COSMO-SkyMed high resolution fully polarimetric synthetic aperture radar exploitation - the ‘strip-spot’ approach

This focused study aimed to generate a fully polarimetric synthetic aperture radar (SAR) data set of 1 m resolution based on the spotlight and stripmap COSMO-SkyMed (CSK) satellite data fusion. The results show the feasibility of overcoming the limitation of the nominal 3 m resolution generated by a series of multi-temporal stripmap SAR data observed in all the polarisations. The CSK satellite system does not allow the observation of cross-polar data in the spotlight acquisition mode because only co-polar data are available. In this work, a fully polarimetric scattering matrix is estimated by using two spotlights in the horizontal horizontal (HH) and vertical vertical (VV) polarisations and two stripmaps in the horizontal vertical (HV) and vertical horizontal (VH) polarisations. The stripmaps were resampled and super-resolved by using the amplitude and phase estimation of sinusoids (APES) filter to achieve the spotlight resolution. The results show that the proposed strip-spot approach has immediate operative applications.     

A comparison of TerraSAR-X Quadpol backscattering with RapidEye multispectral vegetation indices over rice fields in the Mekong Delta,Vietnam

Satellite-based multispectral imagery and/or synthetic aperture radar (SAR) data have been widely used for vegetation characterization, plant physiological parameter estimation, crop monitoring or even yield prediction. However, the potential use of satellite-based X-band SAR data for these purposes is not fully understood. A new generation of X-band radar satellite sensors offers high spatial resolution images with different polarizations and, therefore, constitutes a valuable information source. In this study, we utilized a TerraSAR-X satellite scene recorded during a short experimental phase when the sensor was running in full polarimetric ‘Quadpol’ mode. The radar backscatter signals were compared with a RapidEye reference data set to investigate the potential relationship of TerraSAR-X backscatter signals to multispectral vegetation indices and to quantify the benefits of TerraSAR-X Quadpol data over standard dual- or single-polarization modes. The satellite scenes used cover parts of the Mekong Delta, the rice bowl of Vietnam, one of the major rice exporters in the world and one of the regions most vulnerable to climate change. The use of radar imagery is especially advantageous over optical data in tropical regions because the availability of cloudless optical data sets may be limited to only a few days per year. We found no significant correlations between radar backscatter and optical vegetation indices in pixel-based comparisons. VV and cross-polarized images showed significant correlations with combined spectral indices, the modified chlorophyll absorption ratio index/second modified triangular vegetation index (MCARI/MTVI2) and transformed chlorophyll absorption in reflectance index/optimized soil-adjusted vegetation index (TCARI/OSAVI), when compared on an object basis. No correlations between radar backscattering at any polarization and the normalized difference vegetation index (NDVI) were observed.     

Radarsat-2/SAR和MODIS数据联合反演黄土高原地区植被覆盖下土壤水分研究

以黄土高原半干旱区定西为试验区,利用Radarsat-2/SAR和MODIS数据,将由MODIS NDVI估算的植被含水量(VWC)应用到微波散射Water-Cloud模型中校正植被的影响。采用交叉极化(VV/VH)组合方案对植被覆盖下土壤水分的反演进行初步探讨,结果表明:在植被影响校正前,模型反演土壤水分值出现明显低估现象;校正植被影响后,相关系数R由0.13提高到0.44,且通过α=0.01的显著性检验,标准差SD由5.02降低到4.30,有效提高了模型反演土壤水分的准确度。卫星反演的研究区土壤含水量大部分介于10%~30%之间,与实地考察情况一致,较好地反映出区域土壤湿度分布信息。表明,光学和微波协同遥感反演对于提高农田土壤水分遥感反演精度具有较大的应用潜力。     

Snow depth in Dhundi: an estimate based on weighted bias corrected differential phase observations of dual polarimetric bi-temporal Sentinel-1 data

ABSTRACT

The studies on snow depth comprise a crucial area of research in the Indian Himalayas, where the seasonal snow cover primarily drives the rivers and significant water resources. In this paper, the initial estimates of the line of sight displacement obtained using differential interferometric phase in VV and VH polarizations using Sentinel-1 bi-temporal dual polarimetric SAR data corresponding to snow covered and snow free land cover, are improved by applying bias corrections for the snow phase and for residual errors in displacement derived from the corrected snow phase. The bias for the snow phase is computed from the observed phase in VV and VH polarizations for the snow free area and the bias for the residual errors is computed by observing the stationary pixels identified in the snow free area using a digital elevation model. The snow depth is computed as a weighted sum of the corrected displacements in the VV and VH polarization, with the weights derived using the local incidence angle. The snow depth results based on the proposed approach was evaluated with respect to field measurements and a coefficient of determination of 0.628 was observed with an improvement of ~0.4 as compared to the displacement observed in the VV and VH channel using the conventional method.     

基于Sentinel-2的UNVI植被指数及性能对比研究

作物精准识别和分类是农业遥感检测的重要内容,对作物长势监测以及估产十分重要。以美国混合农业带为研究区,基于Sentinel-2时间序列影像,根据其传感器响应函数计算了针对Sentinel-2的通用归一化植被指数（Universal Normalized Vegetation Index,UNVI）,并通过两个对比实验,分析UNVI等6个指数在作物精准分类中的性能。实验一以JM（Jeffries-Matusita）距离为指标对不同作物类别之间的可分性进行分析,结果表明UNVI优于NDVI、EVI、WDRVI、NDre1和NDWI指数,在玉米和棉花、玉米和水稻、玉米和水稻的区分上,UNVI优于其他指数区分能力相当,但在其余的作物组合上如棉花和水稻,NDVI等指数则无法将其很好的区分,此时UNVI指数依然可以表现出较好的区分能力;实验二对6种时间序列指数特征分别使用随机森林和支持向量机进行作物分类,结果表明UNVI指数的总体精度和Kappa系数最高,其次是NDre1指数和WDRVI指数,EVI的总体精度和Kappa系数最低,这表明UNVI比其他6个指数更好地区分了研究区大豆、玉米、棉花和水稻等4种主要作物。综上,基于Sentinel-2时间序列的UNVI指数在进行作物分类时与其他5种遥感植被指数相比,具有较大的优势,UNVI可为农作物长势分析和作物估产研究等农业研究和应用的可选植被指数。