Browsing by Author "Jochem Verrelst"

Now showing 1 - 4 of 4

Crop type discrimination using Geo-Stat Endmember extraction and machine learning algorithms
(Elsevier Ltd, 2024) Prachi Singh; Prashant K. Srivastava; Dharambhai Shah; Manish K. Pandey; Akash Anand; Rajendra Prasad; Rucha Dave; Jochem Verrelst; Bimal K. Bhattacharya; A.S. Raghubanshi
The identification of crop diversity in today's world is very crucial to ensure adaptation of the crop with changing climate for better productivity as well as food security. Towards this, Hyperspectral Remote Sensing (HRS) is an efficient technique that offers the opportunity to discriminate crop types based on morphological as well as physiological features due to availability of contiguous spectral bands. The current work utilized the benefits of Airborne Visible Infrared Imaging Spectrometer- New Generation (AVIRIS-NG) data and explored the techniques for classification and identification of crop types. The endmembers were identified using the Geo-Stat Endmember Extraction (GSEE) algorithm for pure pixels identification and to generate the spectral library of the different crop types. Spectral feature comparison was done among AVIRIS-NG, Analytical Spectral Device (ASD)-Spectroradiometer and Continuum Removed (CR) spectra. The best-fit spectra obtained with the Reference ASD-Spectroradiometer and Pure Pixel spectral library were then used for crop discrimination using the ten supervised classifiers namely Spectral Angle Mapper (SAM), Spectral Information Divergence (SID), Support Vector Machine (SVM), Minimum Distance Classifier (MDC), Binary Encoding, deep learning-based Convolution Neural Network (CNN) and different algorithms of Ensemble learning such as Tree Bag, AdaBoost (Adaptive Boosting), Discriminant and RUSBoost (Random Under Sampling). In total, nine crop types were identified, namely, wheat, maize, tobacco, sorghum, linseed, castor, pigeon pea, fennel and chickpea. The performance evaluation of the classifiers was made using various metrics like Overall Accuracy, Kappa Coefficient, Precision, Recall and F1 score. The classifier 2D-CNN was found to be the best with Overall Accuracy, Kappa Coefficient, Precision, Recall and F1 score values of 89.065 %, 0.871, 87.565%, 89.541% and 88.678% respectively. The output of this work can be utilized for large scale mapping of crop types at the species level in a short interval of time with high accuracy. © 2022 COSPAR
Denoising AVIRIS-NG Data for Generation of New Chlorophyll Indices
(Institute of Electrical and Electronics Engineers Inc., 2021) Prachi Singh; Prashant K. Srivastava; Ramandeep Kaur M. Malhi; Sumit K. Chaudhary; Jochem Verrelst; Bimal K. Bhattacharya; Akhilesh S. Raghubanshi
The availability of Airborne Visible and Infrared Imaging Spectrometer-Next Generation (AVIRIS-NG) data has enormous possibilities for quantification of Leaf Chlorophyll Content (LCC). The present study used the AVIRIS-NG campaign site of Western India for generation and validation of new chlorophyll indices by denoising the AVIRIS-NG data. For validation, concurrent to AVIRIS-NG flight overpass, field samplings were performed. The acquired AVIRIS-NG was subjected to Spectral Angle Mapper (SAM) classifier for discriminating the crop types. Three smoothing techniques i.e., Fast-Fourier Transform (FFT), Mean and Savitzky-Golay filters were evaluated for their denoising capability. Raw and filtered data was used for developing new chlorophyll indices by optimizing AVIRIS-NG bands using VIs based on parametric regression algorithms. In total, 20 chlorophyll indices and corresponding 20 models were developed for mapping LCC in the area. SAM identified 17 crop types in the area, while FFT found to be the best for filtering. Performance of these models when checked based on Pearson correlation coefficient ( {r} ) and Centered Root Mean Square Difference (CRMSD), indicated that LCC-CCI10 based on normalized difference type index formed through Near Infrared band and blue band is the best estimator of LCC ({r}_{textit {cal}}=0.73,{r}_{textit {val}}=0.66,CRMSD=4.97). The approach was also tested using AVIRIS-NG image of the year 2018, which also showed a promising correlation ( {r} =0.704 , CRSMD = 8.98, Bias = -0.5) between modeled and field LCC. © 2001-2012 IEEE.
High resolution retrieval of leaf chlorophyll content over Himalayan pine forest using Visible/IR sensors mounted on UAV and radiative transfer model
(Elsevier B.V., 2023) Prachi Singh; Prashant K. Srivastava; Jochem Verrelst; R.K. Mall; Juan Pablo Rivera; Vikas Dugesar; Rajendra Prasad
Forests play an essential role towards net primary productivity, biological cycles and provide habitat to flora & fauna. To monitor key physiological activities in forest canopies such as photosynthesis, respiration, transpiration, spatially-explicit and precise information of the biochemical (biological) variables such as Leaf Chlorophyll Content (LCC) is required. While lookup-table (LUT)-based Radiative Transfer Model (RTM) inversion against optical remote sensing imagery is regarded as a physically sound and robust approach for retrieving biochemical and biophysical variables, regularization procedures are required to offset the problem of ill-posedness. To optimize the RTM inversion of LCC over a sub-tropical pine forest plantation, in the Western Himalaya, we investigated the role of: (1) cost functions (CFs), (2) added noise, and (3) multiple finest solutions in LUT inversion. Principal CFs were evaluated belonging to three categories: information measures, M-estimates, and minimal contrast approaches. The inversion approaches were applied to a LUT produced by the coupled leaf-canopy model known as PROSAIL RTM and tested in contrast field spectral data obtained from reflectance data derived from UAV (Unmanned Aerial Vehicle) images taken over the canopies of covered pine forests. The Bhattacharyya divergence, an information measure, outperformed all other CFs in LCC inversion, with R2 of 0.94, RMSE of 6.20 μg/cm2 and NRMSE of 12.27% during the validation. The optimized inversion strategy was subsequently applied to a UAV-acquired multispectral image at an 8.2 cm pixel resolution for detailed landscape forest LCC mapping. The associated residuals as provided by the LUT-based inversion provided insights in the spatial consistency of the LCC map. © 2023 Elsevier B.V.
Retrieval of crop traits using PROSAIL-based hybrid radiative transfer model and EnMAP hyperspectral data
(Elsevier B.V., 2025) Prachi Singh; Prashant Kumar Srivastava; Prakash Kumar Jha; Jochem Verrelst; Pashupati Nath Singh; Rajendra B. Prasad
Implementing high spectral resolution imaging from the Environmental Mapping and Analysis Program (EnMAP) paved the way for detailed retrieval of agricultural traits for accurate crop monitoring and management. The proposed methodology involves the integration and detailed analysis of Radiative Transfer Modelling (RTM) with an integrated approach of machine learning (ML) and Active Learning (AL) algorithms for the retrieval of the Leaf Chlorophyll Content (LCC), Carotenoids (Car) and Leaf Area index (LAI) of wheat cropland from the continuous three years of the dataset. Reflectance values of leaf were collected using Analytical Spectral Device (ASD) − Spectroradiometer data ranging from 350-2500 nm and EnMAP satellite hyperspectral data extends spectral data range varies between 420 nm to 1000 nm in the visible and near-infrared (VNIR) of EMR region, and from 900 nm to 2450 nm in the shortwave infrared (SWIR) region for crop parameters mapping for a larger spatial area of Varanasi district, Uttar Pradesh, India. The PROSPECT + SAIL (PROSAIL) RTM was employed to simulate spectral (reflectance) data, and fourteen ML algorithms were assessed for implementation into a hybrid model. Kernel Ridge regression (KRR) was combined with Euclidean-based Diversity (EBD) algorithms to retrieve crop characteristics due to their exceptional accuracy and reduced uncertainty. Spectral profiles were further used to train hybrid models using PCA (Principal Component Analysis) feature selection, and combined techniques (ML + AL) were applied to retrieve LCC, Car, and LAI. Afterwards, biophysical and biochemical spatial large-scale estimation were provided through atmospherically corrected, and noise-removed EnMAP hyperspectral data with the help of a trained and tested hybrid (ML + AL) model and validated with the ground-measured datasets. The performance indicators showed significantly very high values of correlation during calibration (LCC = 0.99, Car = 0.74, and LAI = 0.99) and validation (LCC = 0.66, Car = 0.57, and LAI= 0.88). The work showed that the optimized hybrid (KRR + AL) models customized for EnMAP hyperspectral data can efficiently estimate the wheat biophysical and biochemical parameters in near-real time therefore, expanding this workflow to agricultural fields may enable more effective monitoring and management of wheat crops. © 2025 The Author(s)