Knowledge of the location and extent of agricultural fields is required for many applications, including agricultural statistics, environmental monitoring, and administrative policies. Furthermore, many mapping applications, such as object-based classification, crop type distinction, or large-scale yield prediction benefit significantly from the accurate delineation of fields. Still, most existing field maps and observation systems rely on historic administrative maps or labor-intensive field campaigns. These are often expensive to maintain and quickly become outdated, especially in regions of frequently changing agricultural patterns. However, exploiting openly available remote sensing imagery (e.g., from the European Union’s Copernicus programme) may allow for frequent and efficient field mapping with minimal human interaction. We present a new approach to extracting agricultural fields at the sub-pixel level. It consists of boundary detection and a field polygon extraction step based on a newly developed, modified version of the growing snakes active contours model we refer to as graph-based growing contours. This technique is capable of extracting complex networks of boundaries present in agricultural landscapes, and is largely automatic with little supervision required. The whole detection and extraction process is designed to work independently of sensor type, resolution, or wavelength. As a test case, we applied the method to two regions of interest in a study area in the northern Germany using multi-temporal Sentinel-2 imagery. Extracted fields were compared visually and quantitatively to ground reference data. The technique proved reliable in producing polygons closely matching reference data, both in terms of boundary location and statistical proxies such as median field size and total acreage.

Grassland contributes to carbon storage and animal feed production. Its yield is largely determined by the cutting times of grassland. Previous studies have used remote sensing data for grassland biomass estimation, but only a few studies have focused on SAR remote sensing approaches for automatic grassland cutting status detection. Due to the occurrence of multiple cuttings in a year, it is crucial to effectively monitor grassland cutting events in order to achieve accurate biomass estimations of a whole season. In this study, we examined the capabilities of multilayer perceptron neural networks for automatic grassland cutting status detection using SAR imagery. The proposed model inputs are a time series dataset of VV and VH Sentinel-1 C-band SAR and second-order texture metrics (homogeneity, entropy, contrast and dissimilarity). The proposed approach has been successfully tested on a dataset collected from several fields in Germany in 2016, with an overall accuracy of 85.71% for the validation set.

Measuring the predictability and complexity of 2D data (image) series using entropy is an essential tool for evaluation of systems’ irregularity and complexity in remote sensing and geophysical mapping. However, the existing methods have some drawbacks related to their strong dependence on method parameters and image rotation. To overcome these difficulties, this study proposes a new method for estimating two-dimensional neural network entropy (NNetEn2D) for evaluating the regularity or predictability of images using the LogNNet neural network model. The method is based on an algorithm for converting a 2D kernel into a 1D data series followed by NNetEn2D calculation. An artificial test image was created for the study. We demonstrate the advantage of using circular instead of square kernels through comparison of the invariance of the NNetEn2D distribution after image rotation. Highest robustness was observed for circular kernels with a radius of R = 5 and R = 6 pixels, with a NNetEn2D calculation error of no more than 10%, comparable to the distortion of the initial 2D data. The NNetEn2D entropy calculation method has two main geometric parameters (kernel radius and its displacement step), as well as two neural network hyperparameters (number of training epochs and one of six reservoir filling techniques). We evaluated our method on both remote sensing and geophysical mapping images. Remote sensing imagery (Sentinel-2) shows that brightness of the image does not affect results, which helps keep a rather consistent appearance of entropy maps over time without saturation effects being observed. Surfaces with little texture, such as water bodies, have low NNetEn2D values, while urban areas have consistently high values. Application to geophysical mapping of rocks to the northwest of southwest Australia is characterized by low to medium entropy and highlights aspects of the geology. These results indicate the success of NNetEn2D in providing meaningful entropy information for 2D in remote sensing and geophysical applications.

Field mapping and information on agricultural landscapes is of increasing importance for many applications. Monitoring schemes and national cadasters provide a rich source of information but their maintenance and regular updating is costly and labor-intensive. Automatized mapping of fields based on remote sensing imagery may aid in this task and allow for a faster and more regular observation. Although remote sensing has seen extensive use in agricultural research topics, such as plant health monitoring, crop type classification, yield prediction, and irrigation, field delineation and extraction has seen comparatively little research interest. In this study, we present a field boundary detection technique based on deep learning and a variety of image features, and combine it with the graph-based growing contours (GGC) method to extract agricultural fields in a study area in northern Germany. The boundary detection step only requires red, green, and blue (RGB) data and is therefore largely independent of the sensor used. We compare different image features based on color and luminosity information and evaluate their usefulness for the task of field boundary detection. A model based on texture metrics, gradient information, Hessian matrix eigenvalues, and local statistics showed good results with accuracies up to 88.2%, an area under the ROC curve (AUC) of up to 0.94, and F1 score of up to 0.88. The exclusive use of these universal image features may also facilitate transferability to other regions. We further present modifications to the GGC method intended to aid in upscaling of the method through process acceleration with a minimal effect on results. We combined the boundary detection results with the GGC method for field polygon extraction. Results were promising, with the new GGC version performing similarly or better than the original version while experiencing an acceleration of 1.3× to 2.3× on different subsets and input complexities. Further research may explore other applications of the GGC method outside agricultural remote sensing and field extraction.

A growing world population, increasing prosperity in emerging countries, and shifts in energy and food demands necessitate a continuous increase in global agricultural production. Simultaneously, risks of extreme weather events and a slowing productivity growth in recent years has caused concerns about meeting the demands in the future. Crop monitoring and timely yield predictions are an important tool to mitigate risk and ensure food security. A common approach is to combine the temporal simulation of dynamic crop models with a geospatial component by assimilating remote sensing data. To ensure reliable assimilation, handling of uncertainties in both models and the assimilated input data is crucial. Here, we present a new approach for data assimilation using particle swarm optimization (PSO) in combination with statistical distance metrics that allow for flexible handling of model and input uncertainties. We explored the potential of the newly proposed method in a case study by assimilating canopy cover (CC) information, obtained from Sentinel-2 data, into the AquaCrop-OS model to improve winter wheat yield estimation on the pixel-and field-level and compared the performance with two other methods (simple updating and extended Kalman filter). Our results indicate that the performance of the new method is superior to simple updating and similar or better than the extended Kalman filter updating. Furthermore, it was particularly successful in reducing bias in yield estimation.