Figure - available from: Remote Sensing
This content is subject to copyright.
Source publication
National-level mapping of crop types is important to monitor food security, understand environmental conditions, inform optimal use of the landscape, and contribute to agricultural policy. Countries or economic regions currently and increasingly use satellite sensor data for classifying crops over large areas. However, most methods have been based...
Citations
... Crop mapping index-based approaches have simpler structures and easier replicability compared to data-driven algorithms. National-scale crop data products have been successfully generated using knowledge-based approaches (Planque et al., 2021;Zhang et al., 2017). However, it is challenging to discriminate multiple crop types since crop mapping indices are typically designed to highlight one single targeted crop based on its key phenological stages (Ashourloo et al., 2019;Xu et al., 2023). ...
... Differences in plant structure between broad or thin-leafed crops result in different backscatter patterns in polarization, which can enhance crop mapping using SAR data (Lussem et al., 2016). Recent studies suggested that the VH/VV ratio was sensitive to changes in the vegetation structure of several winter crops (e.g., winter wheat) (Nasrallah et al., 2019;Planque et al., 2021). However, similar VH/VV temporal signatures might appear among crops with different morphological structures (Arias et al., 2020). ...
Complex cropping patterns with crop diversity are an underexploited treasure for global food security. However, significant methodological and dataset gaps in fully characterizing cropland cultivated with multiple crops and rotation sequences hinder our ability to understand and promote sustainable agricultural systems. Existing crop mapping models are challenged by the deficiency of ground reference data and the limited transferability capabilities across large spatial domains. This study aimed to fill these gaps by proposing a robust Complex Cropping Pattern Mapping framework (CCPM) capable of national-scale automatic applications using the Sentinel-1 SAR and Sentinel-2 MSI time series datasets. The CCPM framework addresses these challenges by integrating knowledge-based approaches & data-driven algorithms (Dual-driven model) and Phenological Normalization. The CCPM framework was implemented over conterminous China with complex cropping systems dominated by smallholder farms, and the first national-scale 10-m Cropping pattern map with descriptions of cropping intensity and 10 crops in China (ChinaCP-T10) in 2020 was produced. The efficiency of the CCPM framework was validated when evaluated by 18,706 ground-truth reference datasets, with an overall accuracy of 91.47 %. Comparisons with existing crop data products revealed that the ChinaCP-T10 offered more comprehensive and consistent information on diverse cropping patterns. Dominant cropping patterns diversified from single maize in northern China, winter wheat-maize in North China Plain, single oilseeds in Western China, to single rice or double rice in Southern China. The key cropping patterns changed from double-grain cropping, single grain to single cash cropping with increasing altitudes. There were 151,744 km 2 planted areas of double grain cropping patterns in China, and multiple cropping accounted for 36.1 % of grain cultivated area nationally. Over 80 % of grain production was mainly implemented at lower altitudes as the Non-Grain Production (NGP) ratio enhanced from 32 % within elevations below 200 m to over 70 % among elevations above 700 m. Consistent datasets on complex cropping patterns are essential, given the significant roles of diversification and crop rotations in sustainable agriculture and the frequently observed inconsistencies in existing crop data products based on thematic mapping.
... Over the past few decades, a lot of work has been done in agricultural classification by utilizing the Sentinel-2 satellite dataset Simón Sánchez et al. 2022). Studies utilizing data from optical satellites such as Sentinel-2 demonstrated significant effectiveness for agricultural classification (Yang et al. 2024), whereas microwave sensors-based satellites such as Sentinel-1 show promising results in all-weather capabilities for depicting surface information (Planque et al. 2021). Each of the sensors has its own advantages and drawbacks. ...
Agriculture is crucial for economic growth, rural development, and food security. Remote sensing aids in cost-effective agricultural mapping, but challenges like limited resolution, atmospheric errors, and cloud interference in satellite imagery hinder accurate monitoring. To overcome these challenges, this study introduces an image fusion-based framework for detecting agricultural changes with the incorporation of optical and microwave satellite data. It integrates the strengths of multi-source sensors to provide enhanced accuracy in classification and change detection procedures, especially in the detection of agricultural variation. The novelty of the work lies within the incorporation of a fusion process in the post-classification-based change detection method and the newly developed framework named fusion-based post-classification change detection (FPCD). To perform the in-depth analysis of FPCD, various fusion methods i.e., (a) Gram-Schmidt (GS) and (b) PC spectral (PCS), and various machine learning-based classification methods i.e., (a) maximum likelihood classifier (MLC), (b) minimum distance classifier (MDC), and (c) support vector machine (SVM) was utilized. Imagery from the Sentinel-1 with VH and VV polarization bands and Sentinel-2 L2A (Level 2A) satellites were acquired over a region in Punjab, India, known for its fertile soil and significant contribution to wheat production. The experimental outcomes confirmed the effectiveness of the proposed FPCD framework by more than overall accuracy (96.5% and 92.74% in classified and change maps, respectively) as compared to other existing frameworks i.e., SVM-based (94.7% and 88.29% in classified and change maps, respectively) and MDC (88% and 81.19% in classified and change maps, respectively). These outcomes are satisfactory enough to monitor multitemporal agricultural variations at a large scale in an effective manner.
... Consequently, time series of remote sensing images have become increasingly available. Regional land cover maps with a spatial resolution of 30 m have been generated [3,[6][7][8][9][10][11][12][13] through time series analysis of satellite images. Furthermore, the authors of [14] developed a land cover mapping method (LCMM) using multi-classifiers and multisource remotely sensed imagery, employing HJ-1/CCD and Landsat time series images to produce finer land cover maps at Heihe River Basin (HRB) for the years 1986,1990,1995,2000,2005, and 2011-2015 [14,15]. ...
Long time series of annual land cover with fine spatio-temporal resolutions play a crucial role in studying environmental climate change, biophysical modeling, carbon cycling models, and land management. Despite a strong consistency exhibited by several publicly available medium to fine resolution global land cover datasets, significant discrepancies exist at the regional scale; moreover, only every 5/10 year land cover were available. Consequently, high-quality annual land cover datasets before 2000 are unavailable in China. In this study, we proposed a deep learning-based method by integrating multiple remote sensing data from different platforms with historical high spatial resolution land cover datasets (CNLUCC) to derive the 30 m annual land cover maps from 1980 to 1990 for Qilian Mountain. First, the super-resolution generative adversarial network models for upscaling the 5.5 km AVHRR NDVI to 250 m were established by employing the AVHRR and MODIS NDVI data with the same year as input, and the early time series AVHRR NDVI data were subsequently upscaled to 250 m through the above models. Second, the breaks for the additive seasonal and trend (BFAST) change detection algorithm was applied to the upscaled time series NDVI data to detect the change time of different land cover types. Third, the CNLUCC data in 1980 and 1990 were updated to annual land cover datasets from 1980 to 1990 and the annual mapping results provided insights into the dynamic processes of urbanization, deforestation, water bodies, and farmland from 1980 to 1990. Finally, comprehensive analysis and validation were carried out for evaluation and an overall accuracy of 77.26% for the land cover product in 1986 was achieved.
... Sentinel-1 SAR transmits and receives within the C-band microwave region, where interactions are affected by the amount and structure of plant material and ground conditions, such as moisture and surface roughness (Veloso et al., 2017). The vegetation influences the backscatter of VV and VH polarizations, with VV found to be more sensitive to ground conditions (Planque et al., 2021). Freely available multi-date Sentinel-1 SAR data (Spatial Resolution 20m (10*10), Temporal Resolution 12 days, VV polarization, Ground Range Detection Product, high resolution) from 19 May to 15 November for two consecutive years (2019 & 2020) was used for kharif rice classification in the study districts (https://scihub.copernicus.eu/dhus/). ...
Rice is a major staple crop in the world, and therefore, accurate acreage assessment of crop area is essential. Traditional methods of crop acreage estimation often rely on ground surveys and manual data collection, which are labor intensive, time-consuming, and may lack spatial coverage. In recent years, the availability of satellite remote sensing data has revolutionized the monitoring of agricultural systems by providing frequent, spatially extensive, and cost-effective information. This study presents a remote sensing approach for assessing crop area and identifying sowing dates using satellite data. We used the Sentinel-1 synthetic aperture radar (SAR) data to estimate the acreage of rice crop and sowing/transplanting of kharif rice during the years 2019 & 2020. Results showed that the rice area and sowing dates were validated using ground truth, and an overall accuracy of 87 to 92% was achieved for both years. Satellite based assessment of crop area and sowing date identification provides reliable information with high spatial and temporal resolution, enabling timely decision-making for farmers, policymakers, and agricultural stakeholders. Furthermore, the scalability and cost-effectiveness of satellite remote sensing make it a valuable tool for large-scale agricultural monitoring and management. In conclusion, this study highlights the potential of Sentinel-1 SAR satellite data for early-stage assessment of rice crop areas, offering a practical and efficient solution for monitoring agricultural dynamics at regional and global scales. Continued advancements in satellite technology and data processing techniques promise further improvements in the accuracy and applicability of remote sensing-based approaches for agricultural monitoring and management.
... In this context, several approaches have been developed to explore radar measurements to monitor vegetation dynamics, especially with the increasing availability of C-band SAR data with high resolution and repetitiveness after the launch of Sentinel-1A/B in 2014 and 2016, respectively (Allies et al. 2021;Bell et al. 2020;Kussul et al. 2017;Rembold et al. 2015). Various studies have relied on radiometric information by calculating the VH/VV cross-polarization ratio (Amal et al. 2021;Planque et al. 2021;Veloso et al. 2017;Vreugdenhil et al. 2018;Yunjin and Van Zyl 2009), the radar vegetation index (RVI) (Yunjin and Van Zyl 2009) and its truncated RVI (Dipankar et al. 2020;Haldar et al. 2022). Owing to the potential of Sentinel-1 data to monitor the vegetation dynamics, Song et al. (2021) highlighted the contribution of the C-band crosspolarization ratio in fusion with optical data to improve the crop type classification. ...
Annual crop monitoring is a key parameter for managing agricultural strategies. Several studies have relied on remote sensing products such as the normalized difference vegetation index (NDVI) as a vegetation dynamic metric. However, the dependence of optical data on weather conditions limits its availability. In this study, we reconstruct the NDVI time series of wheat fields using the moving averages of the Sentinel-1 normalized VH/VV cross-polarization ratio (IN) and the interferometric coherence in VV polarization over wheat selected fields in a semiarid site in Tunisia during two seasons, from 2018 to 2020. The crop cycle is divided into two periods: before and after the heading phase, which occurs in approximately the middle of March. Due to the volume-scattering impact, the second phase is divided into the ripening and maturation phase (NDVI ≥0.4) and senescence phase (NDVI <0.4). To estimate the NDVI values, different methods are used: curve-fitting equations and machine learning regressors such as the random forest (RF) and the support vector regressor (SVR). Low root mean square error (RMSE) values characterize NDVI estimation during the first period. In the second period, the RMSE values reach 0.06 when the NDVI is lower than 0.4. When the NDVI values exceed 0.4 in the second period, lower accuracy marks the NDVI estimation using the curve-fitting equations as a function of IN or coherence. Relative low accuracy characterizes the regression algorithms’ estimations when NDVI ≥ 0.4 compared to their performance during the aforementioned periods. The proposed approach was tested on different wheat fields. The NDVI estimations are characterized by RMSE values varying between 0.12 and 0.19. The use of RF and SVR outperformed the curve-fitting methods with an RMSE equal to 0.12. The present findings revealed the high accuracy of the proposed approach to estimate the missing values of wheat fields NDVI values during the vegetation development period until heading and the senescence phase. The presence of the mutual effect of the vegetation water content and its volume complicated the NDVI estimation using the C-band data.
... The decline in accuracy was attributed to the change in the crop structure and, thus, the reflectance of different cover types. Reflectance and the selection of particular bands to improve classification accuracy have been addressed in different studies (Peña et al., 2017;Al-Shammari et al., 2020;Kobayashi et al., 2020;Planque et al., 2021). In this study, the poor crop structure (caused by poor crop density) in the 2018 growing season reduced the reflected EVI to a median of ∼0.32 for cereal and of 0.52 for canola in August (mid-season), whereas the EVI values in 2017 and 2019 had increased to higher than 0.50 for both crop types (Fig. 4. B). ...
... The difference between crop types in terms of structure and during phenological stages has an influence on the backscattering intensities (Forkuor et al., 2014). This is because the VV is influenced by the volume scattering and VH is influenced by the double-bouncing scattering between stem and ground (Planque et al., 2021). Therefore, the high importance of the VV and VH predictors proved that these two predictors can discriminate different crop types with different structures. ...
... Further, authors achieved the KA of 0.7998 and 0.7499 for VH and VV polarization respectively. Ref. [43] parcel-based classification approach and well-versed by parallel analysis of knowledge-based crop growth stages and Sentinel-1 C-band SAR time series data. The authors achieved OA ranging between 85.8% and 90.6%. ...
With the growing popularity of deep learning, semantic segmentation using convolutional neural networks (CNNs) has proven the state of the art in the pixel-level classification of the remote sensed multi-temporal images captured by satellites such as Sentinel-1A, Sentinel-1B, Sentinel-2, and Landsat-8. Among these, the temporal Sentinel-1B data has widely been used for crop mapping. This research is entirely focused on crop classification based on Sentinel-1B synthetic aperture radar imagery. We have implemented seven popular CNN-based deep learning models and their variations for the segmentation and classification of the pre-processed Sentinel-1B SAR images. Further, we proposed an approach by collaborating the UNet and SEResNext50 as the backbone along with the custom loss function (a hybrid of dice loss and focal loss) and evaluated its performance qualitatively and quantitatively using various metrics. It is observed that the proposed approach is able to achieve an average IoU of 0.6465, average precision of 0.7371, average recall of 0.7191, and average F1-score of 0.7352. Based on the per-pixel confusion matrix the proposed approach achieves an overall accuracy of 98.69% and a kappa coefficient of 0.87. Further, the applicability in the context of Indian agriculture, as well as the current assistance provided by the Mahalanobis National Crop Forecast Centre as part of the Forecasting Agricultural output using Space, Agrometeorology, and Land-based observations programme has been discussed. We have also suggested a few proposals that can be considered by the Ministry of Agricultural and Farmer Welfare, India for the development of the application/platform to provide the ground labels/reference in formats such as GeoTiff or shapefile.
... In recent years, Capsule networks [10] have gained more popularity than CNN in different applications such as plant disease detection, text classification, tumor classification, bioinformatics, and simple classification problems. In contrast to CNN, which encodes information in a scalar manner, capsule networks are groups of neurons that encode and store spatial information in vector form. ...
The identification of plant diseases is one of the most essential and difficult concerns in agriculture, necessitating solutions with a brighter light. With the onset of artificial intelligence (AI), machine learning (ML) and deep learning (DL) algorithms have aided farmers in identifying and classifying plant features with a high degree of intellectual precision. However, accurate disease classification in plants is essential for empowering farmers to cultivate more and produce more. This study therefore presents a unique assembly of attention, capsule, and feedforward classification layers for reaching the maximum classification accuracy for plant diseases. The proposed system uses user-defined customized Convolutional Transfer Learning networks (CTLN) to extract features and the attention networks exclude unnecessary features and highlight only critical features for classification. Finally, the selected characteristics are sent to the Feedforward Capsule networks to improve performance. This paper proposes a paradigm that overcomes the constraints of existing deep learning networks and drastically decreases the computing burden. The suggested network is thoroughly evaluated utilizing Plant Village databases containing over 50,000 photos of healthy and diseased plants. The performance metrics of the proposed method are evaluated and compared to those of other learning networks. Compared to previous models, experimental results indicate that the proposed model has a 99.8 percent accuracy rate, lending support to the new categorization method that benefits farmer well-being.
... The classification results were assessed using the seven most prevalent statistical principles, which were all taken from the classification's confusion matrix. These criteria were producer accuracy (PA) or precision, user accuracy (UA) or recall, omission error (OE = 1 − PA), commission error (CE = 1 − UA), overall accuracy (OA), Kappa coefficient (KC) (Yang et al. 2020;Powers and Ailab 2020;Planque et al. 2021), and mean intersection over union (mIoU) Zheng et al. 2022). ...
Precise and cost-effective mapping of crop and other land cover (LC) types is essential for food security, precision agriculture, and water distribution management. However, accurate identification of the extent of crop and LC types is still a challenge. However, the efficiency of machine learning (ML) and deep learning (DL) algorithms were evaluated and assimilated with time-series satellite information and In-situ data for detecting crop and other LC types. Therefore, in the current study, a framework was established on a deep convolutional neural network (CNN) using Sentinel-2 time-series satellite datasets to identify rice crop and other LC types without survey data. The normalized difference vegetation index (NDVI) stack was developed through Google Earth Engine (GEE) and adopted as input features. Three widely used ML algorithms, random forest (RF), support vector machine (SVM) and classification and regression trees (CART), and four DL models (Swin Transformer (ST), HRNet, two-dimensional convolutional neural network (2D-CNN) and long short-term memory (LSTM) were employed to perform the classification of rice crop and four non-crop classes. The overall accuracies were determined to be 93.95%, 91.67%, 89.80%, 86.89%, 81.01%, 76.51%, and 72.9% and the kappa coefficients were 92.35%, 89.49%, 87.13%, 83.52%, 76.12%, 70.56%, and 66.56%, of corresponding methods respectively. The findings showed that DL methods outperformed ML methods, while ST methods yielded the highest accuracy, and input data can be prepared for LULC classification without ground survey. Accurate classification of LC types and timely assessment of rice crop estimation can provide valuable statistics for state officials, decision-makers, and planners.
... RS is widely used in precision agricultural (PA) applications as it is considered a reliable source for extracting phenological information about crops [7][8][9][10]. The availability of high spectral, spatial, and temporal RS data, including multi-spectral [11,12], hyperspectral [13], and synthetic aperture radar (SAR) [6,[14][15][16][17] have opened new possibilities in crop-type mapping [3,[18][19][20][21], crop health [22] and yield estimation [4,[23][24][25]. In the early 21st century, Landsat and moderate resolution imaging spectroradiometer (MODIS) multi-spectral data were relied on for crop types [26]. ...
Climate change and the COVID-19 pandemic have disrupted the food supply chain across the globe and adversely affected food security. Early estimation of staple crops can assist relevant government agencies to take timely actions for ensuring food security. Reliable crop type maps can play an essential role in monitoring crops, estimating yields, and maintaining smooth food supplies. However, these maps are not available for developing countries until crops have matured and are about to be harvested. The use of remote sensing for accurate crop-type mapping in the first few weeks of sowing remains challenging. Smallholder farming systems and diverse crop types further complicate the challenge. For this study, a ground-based survey is carried out to map fields by recording the coordinates and planted crops in respective fields. The time-series images of the mapped fields are acquired from the Sentinel-2 satellite. A deep learning-based long short-term memory network is used for the accurate mapping of crops at an early growth stage. Results show that staple crops, including rice, wheat, and sugarcane, are classified with 93.77% accuracy as early as the first four weeks of sowing. The proposed method can be applied on a large scale to effectively map crop types for smallholder farms at an early stage, allowing the authorities to plan a seamless availability of food.
