Leaf chlorophyll content retrieval from airborne hyperspectral remote sensing imagery

Hyperspectral remote sensing has great potential for accurate retrieval of forest biochemical parameters. In this paper, a hyperspectral remote sensing algorithm is developed to retrieve total leaf chlorophyll content for both open spruce and closed forests, and tested for open forest canopies. Ten black spruce (Picea mariana (Mill.)) stands near Sudbury, Ontario, Canada, were selected as study sites, where extensive field and laboratory measurements were carried out to collect forest structural parameters, needle and forest background optical properties, and needle biophysical parameters and biochemical contents chlorophyll a and b. Airborne hyperspectral remote sensing imagery was acquired, within one week of ground measurements, by the Compact Airborne Spectrographic Imager (CASI) in a hyperspectral mode, with 72 bands and half bandwidth 4.25-4.36 nm in the visible and near-infrared region and a 2 m spatial resolution. The geometrical-optical model 4-Scale and the modified leaf optical model PROSPECT were combined to estimate leaf chlorophyll content from the CASI imagery. Forest canopy reflectance was first estimated with the measured leaf reflectance and transmittance spectra, forest background reflectance, CASI acquisition parameters, and a set of stand parameters as inputs to 4-Scale. The estimated canopy reflectance agrees well with the CASI measured reflectance in the chlorophyll absorption sensitive regions, with discrepancies of 0.06%-1.07% and 0.36%-1.63%, respectively, in the average reflectances of the red and red-edge region. A look-up-table approach was developed to provide the probabilities of viewing the sunlit foliage and background, and to determine a spectral multiple scattering factor as functions of leaf area index, view zenith angle, and solar zenith angle. With the look-up tables, the 4-Scale model was inverted to estimate leaf reflectance spectra from hyperspectral remote sensing imagery. Good agreements were obtained between the inverted and measured leaf reflectance spectra across the visible and near-infrared region, with R² = 0.89 to R² = 0.97 and discrepancies of 0.02%-3.63% and 0.24%-7.88% in the average red and red-edge reflectances, respectively. Leaf chlorophyll content was estimated from the retrieved leaf reflectance spectra using the modified PROSPECT inversion model, with R² = 0.47, RMSE = 4.34 μg/cm², and jackknifed RMSE of 5.69 μg/cm² for needle chlorophyll content ranging from 24.9 μg/cm² to 37.6 μg/cm². The estimates were also assessed at leaf and canopy scales using chlorophyll spectral indices TCARI/OSAVI and MTCI. An empirical relationship of simple ratio derived from the CASI imagery to the ground-measured leaf area index was developed (R² = 0.88) to map leaf area index. Canopy chlorophyll content per unit ground surface area was then estimated, based on the spatial distributions of leaf chlorophyll content per unit leaf area and the leaf area index.     

Estimation of yellow starthistle abundance through CASI-2 hyperspectral imagery using linear spectral mixture models

The invasive weed yellow starthistle (Centaurea solstitialis) has infested between 4 and 6 million hectares in California. It often forms dense infestations and rapidly depletes soil moisture, preventing the establishment of other species. Precise assessment of its canopy cover, especially low-density abundance in the earlier growing season, is the key to effective management. Compact Airborne Spectrographic Imager 2 (CASI-2) hyperspectral imagery was acquired at the western edge of California's Central Valley grasslands on July 15, 2003. Four linear spectral mixture models (LSMM) were investigated from the original CASI-2 data. Band selections based upon residual analysis and feature extraction (PCA) were further explored to reduce the data dimension. All approaches, except four band-selection unconstrained LSMMs, provide consistent results. The uncertainty of the PCA-based LSMM was estimated through a Monte-Carlo simulation. The maximum standard deviation was approximately 11%. The results suggest that unmixing CASI-2 imagery could be used for estimating and mapping yellow starthistle for larger regional areas.     

谐波分析光谱角制图高光谱影像分类

目的 针对光谱角制图(SAM)分类算法对高光谱像元光谱曲线的局部特征和其辐射强度不敏感,而且易受噪声和维数灾难影响,致使分类效率低和精度较差等缺陷,将谐波分析(HA)技术引入到SAM高光谱影像分类中,提出一种基于谐波分析的光谱角制图(HA-SAM)高光谱影像分类算法.方法 利用HA技术将高光谱影像从光谱维变换到能量谱特征维空间,并提取低次谐波分量及特征系数(谐波余项、相位和振幅),用特征系数组成的向量代替光谱向量,对高光谱影像进行SAM分类.结果 将SAM和HA-SAM同时应用于EO-1卫星的Hyperion高光谱影像分类,通过对比和分析,验证了HA-SAM的优越性,再选择AVIRIS(airborne visible infrared imaging spectrometer)高光谱影像对HA-SAM进行验证,结果表明该算法具有较强的普适性.结论 HA-SAM提高了传统SAM高光谱影像分类的效率和精度,而且适用性较强具有良好的应用前景.     

Evaluation of Random Forest and Adaboost tree-based ensemble classification and spectral band selection for ecotope mapping using airborne hyperspectral imagery

Detailed land use/land cover classification at ecotope level is important for environmental evaluation. In this study, we investigate the possibility of using airborne hyperspectral imagery for the classification of ecotopes. In particular, we assess two tree-based ensemble classification algorithms: Adaboost and Random Forest, based on standard classification accuracy, training time and classification stability. Our results show that Adaboost and Random Forest attain almost the same overall accuracy (close to 70%) with less than 1% difference, and both outperform a neural network classifier (63.7%). Random Forest, however, is faster in training and more stable. Both ensemble classifiers are considered effective in dealing with hyperspectral data. Furthermore, two feature selection methods, the out-of-bag strategy and a wrapper approach feature subset selection using the best-first search method are applied. A majority of bands chosen by both methods concentrate between 1.4 and 1.8 μm at the early shortwave infrared region. Our band subset analyses also include the 22 optimal bands between 0.4 and 2.5 μm suggested in Thenkabail et al. [Thenkabail, P.S., Enclona, E.A., Ashton, M.S., and Van Der Meer, B. (2004). Accuracy assessments of hyperspectral waveband performance for vegetation analysis applications. Remote Sensing of Environment, 91, 354-376.] due to similarity of the target classes. All of the three band subsets considered in this study work well with both classifiers as in most cases the overall accuracy dropped only by less than 1%. A subset of 53 bands is created by combining all feature subsets and comparing to using the entire set the overall accuracy is the same with Adaboost, and with Random Forest, a 0.2% improvement. The strategy to use a basket of band selection methods works better. Ecotopes belonging to the tree classes are in general classified better than the grass classes. Small adaptations of the classification scheme are recommended to improve the applicability of remote sensing method for detailed ecotope mapping.     

An airborne lidar sampling strategy to model forest canopy height from Quickbird imagery and GEOBIA

High-resolution digital canopy models derived from airborne lidar data have the ability to provide detailed information on the vertical structure of forests. However, compared to satellite data of similar spatial resolution and extent, the small footprint airborne lidar data required to produce such models remain expensive. In an effort to reduce these costs, the primary objective of this paper is to develop an airborne lidar sampling strategy to model full-scene forest canopy height from optical imagery, lidar transects and Geographic Object-Based Image Analysis (GEOBIA). To achieve this goal, this research focuses on (i) determining appropriate lidar transect features (i.e., location, direction and extent) from an optical scene, (ii) developing a mechanism to model forest canopy height for the full-scene based on a minimum number of lidar transects, and (iii) defining an optimal mean object size (MOS) to accurately model the canopy composition and height distribution. Results show that (i) the transect locations derived from our optimal lidar transect selection algorithm accurately capture the canopy height variability of the entire study area; (ii) our canopy height estimation models have similar performance in two lidar transect directions (i.e., north-south and west-east); (iii) a small lidar extent (17.6% of total size) can achieve similar canopy height estimation accuracies as those modeled from the full lidar scene; and (iv) different MOS can lead to distinctly different canopy height results. By comparing the best canopy height estimate with the full lidar canopy height data, we obtained average estimation errors of 6.0 m and 6.8 m for conifer and deciduous forests at the individual tree crown/small tree cluster level, and an area weighted combined error of 6.2 m, which is lower than the provincial forest inventory height class interval (i.e., ≈ 9.0 m).     

Kernel-based regularized-angle spectral matching for target detection in hyperspectral imagery

Target detection is one of the most important applications of hyperspectral imagery in the field of both civilian and military. In this letter, we firstly propose a new spectral matching method for target detection in hyperspectral imagery, which utilizes a pre-whitening procedure and defines a regularized spectral angle between the spectra of the test sample and the targets. The regularized spectral angle, which possesses explicit geometric sense in multidimensional spectral vector space, indicates a measure to make the target detection more effective. Furthermore Kernel realization of the Angle-Regularized Spectral Matching (KAR-SM, based on kernel mapping) improves detection even more. To demonstrate the detection performance of the proposed method and its kernel version, experiments are conducted on real hyperspectral images. The experimental tests show that the proposed detector outperforms the conventional spectral matched filter and its kernel version.     

Detection and classification of invasive saltcedar through high spatial resolution airborne hyperspectral imagery

We evaluated the performance of airborne HyperSpecTIR (HST) images for detecting and classifying the invasive riparian vegetation saltcedar along the Muddy River in Clark County, Nevada. HyperSpecTIR image reflectance spectra (227 bands, 450–2450 nm) were acquired for the following four vegetation covers: invasive saltcedar, native honey mesquite, grassland patches and crops. We compared five feature reduction approaches: band selection based on Jeffreys–Matusita distance, principal component analysis (PCA), minimum noise fraction (MNF), segmented principal component transform (SPCT) and segmented minimum noise fraction (SMNF). In addition, maximum likelihood (ML) and two spectral angle mapper (SAM) classifiers were applied to all extracted bands or features. Classification accuracies were compared among all classification approaches. Although the overall accuracy of maximal likelihood classifiers generally surpassed that of SAM classifiers, the highest overall accuracy was achieved by a SMNF-SAM combination with adjusted angular thresholds for classes. We concluded that high spectral and spatial resolution imagery can be used to detect and classify invasive saltcedar in this arid area.     

A Dempster Shafer-based fuzzy multisensor fusion system using airborne lidar and hyperspectral imagery

ABSTRACT

Nowadays, accurate spectral reflectance information is provided by hyperspectral (HS) data while light detection and ranging (lidar) data provides precise information about the height and geometrical properties of the surfaces. In the most research papers, data fusion of disparate sensors significantly improves object classification performance compared to that of just an individual sensor. Previous researches on fusion of these two sensors had problems such as crisp classifiers or simple fuzzy decision-making systems. This article tries to overcome these weaknesses by accurate support vector machine (SVM) and Fuzzy SVM as classifiers in crisp and fuzzy decision fusion system and fusion of two sensors by two different methods based on precise theories of Bayesian and Shafer. Also, the proposed method tries to compare the results of fusion of both data using decision fusion system with stacked features strategy. This study focuses on HS and lidar fusion through three main phases. The first phase is based on the using of Noise Weighted Harsanyi-Farrand-Chang method and principal component analysis to overcome the high dimensionality problem of HS data. The second phase is based on the feature extraction and selection strategy on lidar data. Finally, fuzzy SVM and Dempster Shafer methods are applied as fuzzy classification and fuzzy decision fusion strategies on the feature spaces. A co-registered HS and lidar data set from Houston of U.S.A. by 15 classes was available to examine the effectiveness of the proposed method. The results of this study highlight that the combination of HS and lidar data enable reliable mapping of land cover.     

Comparison of spectral and spatial windows for local anomaly detection in hyperspectral imagery

A new way of implementing two local anomaly detectors in a hyperspectral image is presented in this study. Generally, most local anomaly detector implementations are carried out on the spatial windows of images, because the local area of the image scene is more suitable for a single statistical model than for global data. These detectors are applied by using linear projections. However, these detectors are quite improper if the hyperspectral dataset is adopted as the nonlinear manifolds in spectral space. As multivariate data, the hyperspectral image datasets can be considered to be low-dimensional manifolds embedded in the high-dimensional spectral space. In real environments, the nonlinear spectral mixture occurs more frequently, and these manifolds could be nonlinear. In this case, traditional local anomaly detectors are based on linear projections and cannot distinguish weak anomalies from background data. In this article, local linear manifold learning concepts have been adopted, and anomaly detection algorithms have used spectral space windows with respect to the linear projection. Output performance is determined by comparison between the proposed detectors and the classic spatial local detectors accompanied by the hyperspectral remote-sensing images. The result demonstrates that the effectiveness of the proposed algorithms is promising to improve detection of weak anomalies and to decrease false alarms.     

Improving hyperspectral image classification by combining spectral,texture, and shape features

Several studies have already demonstrated the efficiency of utilizing spatial information in representation and interpretation of hyperspectral (HS) images. Texture and shape features are known as two important categories of spatial information in various applications of image processing. This study tries to utilize texture and shape features extracted from HS images, as well as spectral information, in order to reduce overall classification errors. These features include morphological profiles (MPs), global Gabor features, and features extracted from conventional and segmentation-based grey-level co-occurrence matrices (GLCMs). Various combinations of these spatial features along with spectral information are fed into a support vector machine (SVM) classifier, and the best combinations for different situations are determined. Experiments on the widely used Indian Pines, Pavia University, and Salinas HS data sets demonstrate the efficiency of the proposed framework in comparison with some recent spectral–spatial classification methods.     

Applying image transformation and classification techniques to airborne hyperspectral imagery for mapping Ashe juniper infestations

Ashe juniper (Juniperus ashei Buchholz) in excessive coverage reduces forage production, interferes with livestock management, and degrades watersheds and wildlife habitat on infested rangelands. The objective of this study was to apply minimum noise fraction (MNF) transformation and different classification techniques to airborne hyperspectral imagery for mapping Ashe juniper infestations. Hyperspectral imagery with 98 usable bands covering a spectral range of 475–845 nm was acquired from two Ashe juniper infested sites in central Texas. MNF transformation was applied to the hyperspectral imagery and the transformed imagery with the first 10 and 20 MNF bands was classified using four hard classifiers: minimum distance, Mahalanobis distance, maximum likelihood and spectral angle mapper (SAM). For comparison, the 10‐ and 20‐band MNF imagery was inversely transformed to noise‐reduced 98‐band imagery in the original data space, which was also classified using the four classifiers. Accuracy assessment showed that the first 10 MNF bands were sufficient for distinguishing Ashe juniper from associated plant species (mixed woody species and mixed herbaceous species) and other cover types (bare soil and water). Although the 20‐band MNF imagery provided better results for some classifications, the increase in overall accuracy was not statistically significant. Overall accuracy on the 10‐band MNF imagery varied from 88% for SAM to 93% for minimum distance for site 1 and from 84% for SAM to 94% for maximum likelihood for site 2. The 98‐band imagery derived from the 10‐band MNF imagery resulted in overall accuracy ranging from 91% for both SAM and Mahalanobis distance to 97% for maximum likelihood for site 1 and from 87% for SAM to 93% for minimum distance for site 2. Although both approaches produced comparable classification results, the MNF imagery required smaller storage space and less computing time. These results indicate that airborne hyperspectral imagery incorporated with image transformation and classification techniques can be a useful tool for mapping Ashe juniper infestations.     

Progressive dimensionality reduction by transform for hyperspectral imagery

This paper develops to a new concept, called progressive dimensionality reduction by transform (PDRT), which is particularly designed to perform data dimensionality reduction in terms of progressive information preservation. In order to materialize the PRDT a key issue is to prioritize information contained in each spectral-transformed component so that all the spectral transformed components will be ranked in accordance with their information priorities. In doing so, projection pursuit (PP)-based dimensionality reduction by transform (DRT) techniques are developed for this purpose where the Projection Index (PI) is used to define the direction of interestingness of a PP-transformed component, referred to as projection index component (PIC). The information contained in a PIC is then calculated by the PI and used as the priority score of this particular PIC. Such a resultant PDRT is called progressive dimensionality reduction by projection index-based projection pursuit (PDR-PIPP) which performs PDRT by retaining an appropriate set of PICs for information preservation according to their priorities. Two procedures are further developed to carry out PDR-PIPP in a forward or a backward manner, referred to forward PDR-PIPP (FPDR-PIPP) or backward PDRT (BPDR-PIPP), respectively, where FPDR-PIPP can be considered as progressive band expansion by starting with a minimum number of PICs and adding new PICs progressively according to their reduced priorities as opposed to BPDRT which can be regarded progressive band reduction by beginning with a maximum number of PICs and removing PICs with least priorities progressively. Both procedures are terminated when a stopping rule is satisfied. The advantages of PDR-PIPP allow users to transmit, communicate, process and store data more efficiently and effectively in the sense of retaining data integrity progressively.     

融合高光谱影像三维空谱特征的子空间聚类算法

为提高高光谱遥感影像的聚类精度,将三维空谱特征和子空间聚类算法相结合,提出一种新的稀疏子空间聚类模型,在关注高光谱影像光谱信息的同时也关注了空间上下文信息。首先提取高光谱影像像素点的三种三维空谱特征,然后通过特征对子空间聚类模型的系数矩阵进行加权,使得像素点可被与它最为相似的像素点稀疏表示,从而获得更好的系数矩阵,最后由系数矩阵通过谱聚类获得更好的聚类结果。算法对四个经典高光谱数据集进行实验,并将实验结果与六种聚类算法进行比较,结果表明,所提出的3DF-SSC算法在四个数据集上获得的聚类精度都比其他算法要高,对于同样是利用三维空谱特征的M3DF3△、3DF-SSC算法最高能提高8.62%的精度,而与同样是利用空间上下文信息对子空间聚类算法进行改进的L2-SSC和SS-LRSC算法相比,最高能提高25.18%的精度。     

Estimation of leaf and canopy water content in poplar plantations by means of hyperspectral indices and inverse modeling

This study investigates the applicability of empirical and radiative transfer models to estimate water content at leaf and landscape level. The main goal is to evaluate and compare the accuracy of these two approaches for estimating leaf water content by means of laboratory reflectance/transmittance measurements and for mapping leaf and canopy water content by using airborne Multispectral Infrared and Visible Imaging Spectrometer (MIVIS) data acquired over intensive poplar plantations (Ticino, Italy).At leaf level, we tested the performance of different spectral indices to estimate leaf equivalent water thickness (EWT) and leaf gravimetric water content (GWC) by using inverse ordinary least squares (OLS) regression, and reduced major axis (RMA) regression. The analysis showed that leaf reflectance is related to changes in EWT rather than GWC, with best results obtained by using RMA regression by exploiting the spectral index related to the continuum removed area of the 1200 nm water absorption feature with an explained variance of 61% and prediction error of 6.6%. Moreover, we inverted the PROSPECT leaf radiative transfer model to estimate leaf EWT and GWC and compared the results with those obtained by means of empirical models. The inversion of this model showed that leaf EWT can be successfully estimated with no prior information with mean relative errors of 14% and determination coefficient of 0.65. Inversion of the PROSPECT model showed some difficulties in the simultaneous estimation of leaf EWT and dry matter content, which led to large errors in GWC estimation.At landscape level with MIVIS data, we tested the performance of different spectral indices to estimate canopy water per unit ground area (EWT_canopy). We found a relative error of 20% using a continuum removed spectral index around 1200 nm. Furthermore, we used a model simulation to evaluate the possibility of applying empirical models based on appositely developed MIVIS double ratios to estimate mean leaf EWT at landscape level (). It is shown that combined indices (double ratios) yielded significant results in estimating leaf EWT at landscape level by using MIVIS data (with errors around 2.6%), indicating their potential in reducing the effects of LAI on the recorded signal. The accuracy of the empirical estimation of EWT_canopy and was finally compared with that obtained from inversion of the PROSPECT + SAILH canopy reflectance model to evaluate the potential of both methods for practical applications. A relative error of 27% was found for EWT_canopy and an overestimation of leaf with relative errors around 19%.Results arising from this remote sensing application support the robustness of hyperspectral regression indices for estimating water content at both leaf and landscape level, with lower relative errors compared to those obtained from inversion of leaf and 1D canopy radiative transfer models.     

Endmember search and proportion estimates from airborne hyperspectral surveys

The estimation of areas of land-cover elements is required for many natural resource management programmes and is also used by the mineral and petroleum resource communities either for detection of mineral abundances or monitoring of environmental remediation and other off-site impacts. When the identification of many constituent elements is desired, remote sensors that possess many spectral bands are often deployed, providing data that can be used in a spectroscopic (or other) analysis. At the size of a (remotely sensed) ground sample (represented as an image pixel), which with current technology is typically a few metres, the sample is heterogeneous and typically composed of several biological and geological constituents. It is of interest to first identify the constituent elements and their number and, second, to estimate their relative abundance. When no suitable spectral library is available for a particular data set, an exploratory approach using a blind unmixing method may be used to detect and estimate the endmembers themselves – an exploratory approach because there is no guarantee that the spectral endmembers fitted using blind unmixing will correspond to the ‘pure’ materials of interest to a particular application. Further, if employing a blind unmixing technique to each image in a large multi-image survey independently, there is no guarantee that compatible sets of endmembers will be found to produce maps that are seamless across contiguous images. The aim of this article is to examine the potential for applying blind unmixing at the whole-of-survey level as a way to finding endmembers and proportion maps that are cross-swath consistent and broadscale applicable. We demonstrate that a mosaic of many radiometrically block-adjusted swaths of data from the HyMap airborne hyperspectral imager (HyVista Corporation) can be unmixed as a single image using the Iterated Constrained Endmembers blind unmixing algorithm. The major endmembers are validated against available Analytical Spectral Devices ground spectra and broadscale abundance maps of the type targeted by both vegetation and soil mapping communities are produced.     

Application of spectral features’ ratios for improving classification in partially calibrated hyperspectral imagery: a case study of separating Mediterranean vegetation species

Efficient real-time discrimination of image objects is greatly affected by their radiometry, which is only partly accounted for by image scene calibration. Such calibration treats mainly variations in flux density in the generalized imaged scene plane rather than on the objects’ surface. The proposed methodology uses ratios between secondary parameterizations: e.g., absorption features and spectral derivatives. Clustering in the ratios’ parameter space may allow differentiation between image objects despite limitations regarding their relative calibration. The usefulness of this approach was demonstrated in the challenging task of separating Mediterranean vegetation species using imaging spectroscopy.     

Introducing spectral moment features in analyzing the SpecTex hyperspectral texture database

Hyperspectral imaging provides more information than conventional RGB images. However, its high dimensionality prevents its adaptation to the existing image processing techniques. Defining full-band spectral feature is the first missing step, which is currently dealt with indirectly by band selection or dimension reduction. This article proposes a spectral feature extraction method using the mathematical moments to quantify the shape of the reflectance spectrum from different aspects. A whole family of features is presented by changing the moment attributes. All the features and their combinations are extensively tested in texture analysis of a new hyperspectral image database from textile samples (SpecTex). Two supervised experiments are performed: image patch classification and pixel-wise mosaic image segmentation. The proposed features are compared to four other features: the grayscale intensity, the RGB and CIELab values, and the principal components. Also, three analysis methods are tested: co-occurrence matrix, Gabor filter bank, and local binary pattern. In all cases, the moment features outperformed the opponents. Notably, combining the moment features with complementary attributes remarkably improved the performance. The most discriminative combinations are studied and formulated in this article.     

On combining multiscale deep learning features for the classification of hyperspectral remote sensing imagery

In recent years, satellite imagery has greatly improved in both spatial and spectral resolution. One of the major unsolved problems in highly developed remote sensing imagery is the manual selection and combination of appropriate features according to spectral and spatial properties. Deep learning framework can learn global and robust features from the training data set automatically, and it has achieved state-of-the-art classification accuracies over different image classification tasks. In this study, a technique is proposed which attempts to classify hyperspectral imagery by incorporating deep learning features. Firstly, deep learning features are extracted by multiscale convolutional auto-encoder. Then, based on the learned deep learning features, a logistic regression classifier is trained for classification. Finally, parameters of deep learning framework are analysed and the potential development is introduced. Experiments are conducted on the well-known Pavia data set which is acquired by the reflective optics system imaging spectrometer sensor. It is found that the deep learning-based method provides a more accurate classification result than the traditional ones.     

Assessing growth and yield of wheat using remotely-sensed canopy temperature and spectral indices

Abstract

Prediction models were developed for wheat to assess crop growth in terms of leaf area index, dry matter production and grain yield from remotely-sensed temperature and spectral indices. The cumulative stress degree days (SDD) for the period of flowering to grain formation stage showed significantly higher correlation with dry matter (r= — 0940) and grain yield (r= —0-939) whereas that, for the period grain formation to harvest stage, showed significantly higher correlation lpar;r= —0-967) for crop water use. Significant and positive correlations between dry matter, leaf area and grain yield with infrared/red, normalised difference (ND), transformed vegetation index and greenness index were attained with the latter providing the highest degree of predictability. Spectral indices measured between flowering to milking stages gave the best prediction indicating the suitability of this period for crop growth assessment by this technique. Inter-stage sensitivity analysis by using multiple regression approach also revealed that greenness and transformed vegetation indices could provide better prediction of dry matter and grain yield. From the values of regression coefficients the jointing to beginning of milk formation period of the crop was found to be the most sensitive stage influencing the yield of crop.