page_content stringlengths 1 1.5k | metadata dict |
|---|---|
Sentinel-2 bottom of atmosphere (BOA) reflectance images were used as moderate resolution satellite imagery. Sentinel-2 data products were downloaded from the Copernicus open access hub and imported into a processing platform SNAP toolbox (6.0) provided by European Space Agency (ESA). By using subset command in SNAP, p... | {
"Authors": "Vittorio Mazzia, Lorenzo Comba, Aleem Khaliq, Marcello Chiaberge, Paolo Gay",
"Published": "2020-04-29",
"Summary": "Precision agriculture is considered to be a fundamental approach in pursuing\na low-input, high-efficiency, and sustainable kind of agriculture when\nperforming site-specific manageme... |
Figure 2. (a) Selected test field located in Serralunga d’Alba (Piedmont, northwest of Italy). The boundaries of the three considered parcels, named “Parcel-A”, ”Parcel-B”, and “Parcel-C”, are marked with solid green polygons. The concurrent illustration of low resolution and high-resolution maps derived from satellite... | {
"Authors": "Vittorio Mazzia, Lorenzo Comba, Aleem Khaliq, Marcello Chiaberge, Paolo Gay",
"Published": "2020-04-29",
"Summary": "Precision agriculture is considered to be a fundamental approach in pursuing\na low-input, high-efficiency, and sustainable kind of agriculture when\nperforming site-specific manageme... |
Robotic Monitoring of Colorimetric Leaf Sensors for Precision Agriculture Malakhi Hopkins∗,1 Alice Kate Li∗,1 Shobhita Kramadhati∗,1 Jackson Arnold2 Akhila Mallavarapu1 Chavez Lawrence1 Anish Bhattacharya1 Varun Murali1 Sanjeev J. Koppal2,3 Cherie R. Kagan1 Vijay Kumar1 Abstract—Common remote sensing modalities (RGB, m... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
hyperspectral images. The sensor’s crop healthdependent optical signals can be extracted from the hyperspectral images. The proof-of-concept system is demonstrated in rowcrop environments both indoors and outdoors where it is able to autonomously navigate, locate and obtain a hyperspectral image of all leaf sensors pre... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
typically measure crop or soil reflectance and canopy structure, and the data collected with these methods serve as proxies for crop health traits. More advanced technologies, however, are able to acquire *Equal contribution. 1 GRASP Laboratory, University of Pennsylvania, Pennsylvania, USA 2 University of Florida, Flo... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
leaf sensing technologies are typically bulky, powered electronics that can interfere with crop growth. Here, we co-design a sensor-detector system based on passive, translucent colorimetric, metasurface-based leaf sensors that directly respond to plant stress indicators with a change in their resonance (i.e., color) t... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
of a low-cost, lightweight ground robot platform, BEAST, integrated with a novel hyperspectral imaging system for detecting and measuring deployed leaf sensors; and 3) a computer vision pipeline for online colorimetric leaf sensor detection from RGB images, with visual servoing for targeted hyperspectral reflectance im... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
solution-processible dielectric nanocrystal inks that enable scalable low-cost and high-throughput fabrication. Based on this methodology, we fabricate colorimetric leaf sensors for our purposes of agricultural monitoring. The nanostructured metasurface can be integrated with an adaptive polymer that is responsive to e... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
platforms perform tasks based on visual information gathered about the crop environment. These platforms include the octocopter AGRAS MG-1P [16], Verdant Robotics [17] platforms, and the See & Spray Ultimate from John Deere [18] and [19] Blue River Technologies, used for precise herbicide and pesticide spraying. The ty... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
heavy (16.8kg), but dust and waterproof (IP62), Clearpath Jackal. TerraSentia, developed in [28], was designed specifically for corn stand counting and phenotyping, also weighs a heavy 13.6kg. In our work, we opt for a narrow 38cm form-factor and lightweight 9.0kg ground robot platform, as it provides the stable platfo... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
images and spectra of leaf sensors given the extrinsic calibration between the camera and itself. To solve the task described above, we implement a computer vision pipeline that enables the detector robot to autonomously collect spectral image data from the colorimetric leaf sensors deployed in the environment. Fig. 2 ... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
layer. This design enables high-quality-factor QGM resonances, seen as peaks in the reflectance spectrum, to occur in the visible spectral range at ∼650nm [9]. This ensures that the sensor’s resonances do not overlap with leaves’ green color (∼500-570nm), the high reflectance region in the leaves’ reflectance spectrum ... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
of this work, 2.5cm×2.5cm size optical metasurfaces have been fabricated on glass substrates without the adaptive polymer, with the focus being the first demonstration of the localization of the metasurface sensors by a robotic hyperspectral imaging system and their resonance measurement in outdoor environments. B. Rob... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
a Stereolabs Zed 2i is mounted on the platform. The robot is provided a set of positional waypoints that are sequentially followed using a waypoint controller driving its all-wheel drive motors and a dual-wheel servo steering system. For object detection of the colorimetric leaf sensor, the robot is equipped with a FLI... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
the leaf sensor autonomously, upon detection, during hyperspectral imaging. Finally, the system uses separate batteries for motor and navigation power versus onboard computation and sensing modules, enabling approximately 2 hours of continuous navigation and processing. C. Leaf Sensor Detection and Verification To loca... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
leaf sensor for resonance acquisition. D. Hyperspectral Camera System A hyperspectral camera system is used to measure the colorimetric leaf sensor’s resonance as it provides a comparable resolution (2nm) to lab-based spectrometers. The hyperspectral camera sub-module consists of a pushbroom Headwall NanoHyperspec visi... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
of incidence to the hyperspectral camera lens (when not actuated). Under actuation, the axis of reflection is adjusted to point towards any part of the scene for imaging, enabling foveation, and thus high resolution hyperspectral imagery of a selected area within the FOV. During image capture, the orientation of the mo... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
motorized mirror, we estimate the homography between two planes: (1) the image captured by the RGB camera, and (2) the corresponding hyperspectral image of the same scene. The latter is obtained by merging patches of hyperspectral images captured by moving the motorized mirror in angles of the range [−25◦, 25◦] in both... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
partial cloud coverage, and sunny). A. System Evaluation Metrics To evaluate the performance of our method, component and system-level testing is conducted in three different experimental settings: (1) controlled indoors, (2) unstructured outdoors, and (3) structured outdoors, depicted in Fig. 5. In all three settings,... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
of the model on our evaluation dataset. Metrics. The system-level performance was assessed by the following metrics: (1) Traversal RMSE: measures robot path accuracy, (2-5) Detector mIoU, Precision, Recall: evaluates the YOLO model’s success in correctly identifying leaf sensors, (6, 7) Decision Precision, Recall: eval... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
compared with the extracted reflectance spectrum from a hyperspectral image acquired indoors with BEAST. The resonances are seen as dips in the spectrometer-measured transmission spectrum, and as peaks in the hyperspectral camerameasured reflectance spectrum. Their spectral locations are closely matched and any differe... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
unseen data in various settings: indoors (a), structured outdoors (b), unstructured outdoors in overcast conditions (c), and unstructured outdoors in sunny conditions (d). In all figures, we show the environment, and an enlarged image of the successful YOLO bounding box in the bottom left hand corner. Color values have... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
that the identified sensor’s centroid (x,y) coordinates lie within a specific range in x and y that maximizes the likelihood of successful resonance detection with the required signal-to-noise ratio. If the centroid is is outside this coordinate range, the robot omits this detection. C. Unstructured Outdoor Experimenta... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
one sensor in the first row and one sensor in the third row, and the remaining three included one sensor in the first row and one sensor in the second row. Across the ten outdoor unstructured experiments, Fig. 7: Unstructured environment that the BEAST was evaluated in. Colorimetric leaf sensors were mounted on up to 2... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
required for resonance acquisition. In these ten structured experiments, the system achieved a 100% success rate in acquiring hyperspectral images of the sensor and an 80% success rate in detecting the leaf sensor’s resonance from the acquired hyperspectral images. Fig. 9 shows spectra obtained from the experiments whe... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
this work. First, the utilized hyperspectral camera lens has a Fig. 9: Outdoor characterization of the leaf sensor resonance spectrum TOP using the hyperspectral camera onboard the BEAST during full pipeline testing. Shows the differences in resonance signal-to-noise ratio in varying imaging conditions. BOTTOM Hyperspe... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
metasurface-based leaf sensors with a robotic imaging platform for in-field plant stress monitoring. The leaf sensors are designed to measure plant health more directly than existing remote sensing technologies, as this leaf sensor makes direct contact with the surface of the plant. As the manufacturing cost of these l... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
leaf sensors with a higher signal-to-noise ratio, as well as pursue metasurface designs that exhibit angleindependent resonances. We believe that this will allow us to deploy this system on real farms and further refine our system to handle real-world conditions such as uneven terrain, poor illumination, and obstructed... | {
"Authors": "Malakhi Hopkins, Alice Kate Li, Shobhita Kramadhati, Jackson Arnold, Akhila Mallavarapu, Chavez Lawrence, Varun Murali, Sanjeev J. Koppal, Cherie R. Kagan, Vijay Kumar",
"Published": "2025-07-18",
"Summary": "Common remote sensing modalities (RGB, multispectral, hyperspectral imaging\nor LiDAR) are ... |
XAI-GUIDED ENHANCEMENT OF VEGETATION INDICES FOR CROP MAPPING Hiba Najjar, Francisco Mena∗ University of Kaiserslautern-Landau Kaiserslautern Germany Marlon Nuske, Andreas Dengel German Research Center for Artificial Intelligence Kaiserslautern Germany ABSTRACT Vegetation indices allow to efficiently monitor vegetation... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
in certain cases. 1. INTRODUCTION In recent years, an increasing number of studies have employed Machine Learning (ML) and Deep Learning (DL) techniques to harness remote sensing data for multiple applications related to the sustainable development goals [1]. While such models are proficient in processing raw satellite... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
capability to automatically extract crop-related features and discern interactions between raw bands. To extract scientific insights encoded in the model, eXplainable AI (XAI) techniques can uncover the inner workings of the model, facilitating an understanding of how individual satellite bands contribute to its predic... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
and South Sudan to address this task. The corresponding public datasets used contains satellite image time series captured between January and December 2016 at a 10m resolution, and are labeled with multiple land cover classes [5]. For our study, we merge the two datasets and retain only the pixels corresponding to cro... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
to estimate feature attributions [7]. SVS is grounded in cooperative game theory, which provides a solid theoretical foundation, unlike many other methods [6]. Its robustness has being quantitatively evaluated in the context of a regression task based on time series of satellite data, and has shown superior stability a... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
from S2 data for our analysis, namely the blue (B02), green (B03), red (B04), three RE bands (B05, B06, B07), near-infrared (NIR) (B08), narrow near-infrared (n-NIR) (B8A), and two SWIR (B11, B12) bands. An additional channel, indicating the cloud coverage of the image, is stacked to these bands and used in all our exp... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
attributions We train the GRU-based model using the satellite bands and present the evaluation results on the validation set in the second column of Table 2. This baseline model achieved a score of 67% on both the OA and F1 metrics. In individual classes, high accuracies of 84% and 86% were attained for rice and sorghu... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
x — — — — — — — — — — — — — — NDVI — x — — — — — — x x x — — — — nNDVI — — x — — — — — — — — — — x x NDRE — — — x — — — — x — — x x x — NDRE2 — — — — x — — — — x — — — — — NDRE3 — — — — — x — — — — x — — — — NDMI — — — — — — x — — — — x — — x NDMI2 — — — — — — — x — — — — x — — OA 0.67 0.62 0.62 0.61 0.56 0.51 0.65 0.6... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
NDMI2 (N S2)/(N + S2) This paper channel for the identification of any specific crop. To analyze the results crop-wise, groundnut and soybean highly rely on the first RE band, followed by the red and SWIR1 bands. Sorghum has a similar attribution pattern. Rice has an additional particular dependence on the SWIR2 band. ... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
(NDRE) [20] index that uses the NIR and RE1 bands. We derive two modified indices, NDRE2 and NDRE3, by replacing the first RE channel with the second and third, respectively, to verify whether the relative performance of the three indices align with the attribution of their respective bands. We also incorporate the nor... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
on NDMI, achieving an OA score of 67%. This model outperformed the baseline in identifying three crops: sorghum, groundnut, and yam. The secondbest model is based on the modified version of the same index, NDMI2, which achieved the same class accuracy as the baseline in sorghum and yam, and performed better in soybean.... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
exhibits promising results. The best model based on a single index exhibited an OA 2pp lower than the baseline model, while using two indices achieved a 3pp higher accuracy in the best case. These results highlight the potential of relying solely on one or two VIs for crop identification, especially when carefully sele... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
with NDVI, in contrast to NDRE, which achieved high scores, particularly when combined with NDMI or NDMI2. This observation aligns with the relative average importance of the three RE bands, as shown in Figure 1, suggesting that the first band is more suitable for crop identification. Nonetheless, the second and third ... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
opposite behavior. Overall, one limitation of our XAI-based approach is the reliability of the model. Meaningful explanation results and relevant scientific insights are conditioned by the scientific accuracy of what the model has learned during the training. Since our baseline had an OA score of 67%, we believe that f... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
cases. Overall, our results further indicate that combining two VIs perform better than using a single index, and while some combinations improved the OA over the validation set, an examination of individual crop performance reveals that an index can be highly efficient in identifying certain crops but might struggle w... | {
"Authors": "Hiba Najjar, Francisco Mena, Marlon Nuske, Andreas Dengel",
"Published": "2024-07-11",
"Summary": "Vegetation indices allow to efficiently monitor vegetation growth and\nagricultural activities. Previous generations of satellites were capturing a\nlimited number of spectral bands, and a few expert-d... |
Accepted in 2021 3rd International Conference on Sustainable Technologies for Industry 4.0 (STI), 18-19 December, Dhaka STRIDE-based Cyber Security Threat Modeling for IoT-enabled Precision Agriculture Systems Md. Rashid Al Asif∗, Khondokar Fida Hasan†, Md Zahidul Islam‡, and Rahamatullah Khondoker§ Department of Compu... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
failure would impact severely. Like many other cyberphysical systems, one of the growing challenges to avoid system adversity is to ensure the system’s security, privacy, and trust. But what are the vulnerabilities, threats, and security issues we should consider while deploying precision agriculture? This paper has co... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
sensors, actuators, and devices with the intention of interaction, control, and automated decision-making. Its already been introduced in many countries and rolling in some developing countries like Bangladesh and India to facilitate with lessen human effort, reduced cost, saving time while increasing harvest and profit... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
attack vectors originate from global sites too [4]. Therefore, cyber threats in smart farming management such as precision agriculture are a significant concern for sustainable development that can directly impact crop growth and farmers’ realization. This paper has identified the cyber threat associated with standard pr... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
precision agriculture. Afterward, the research methodology is presented in Section III. Section IV summarizes components, data flows, interactors, and threat models to analyze threats. Section V lists identified cyber security threats plus a recommended list of defense mechanisms against threat categories. The paper is c... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
building efficient and robust system security is crucial. Later on, four-layered smart-agricultural elements are presented along with possible security issues. Finally, they provide a list of security resources (i.e., intrusion/anomaly detection system, firewall, anti-malware, anti-virus, access control, authentication, ... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
followed a systematic and comprehensive approach to maximize security at the component level. So, it might be helpful for a system design and validation process before deployment. A. Omotosho et al. performed threat modeling on eleven IoT health devices based on device assets and access points [8]. They have employed t... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
Threat Modeling Refer to above works, threat modeling resultant details may be helpful in mitigation strategies to increase system security. Moreover, researchers from [10] presented threat modeling on a generic telesurgery system considering its components and data flows. The resultant threat detail might be crucial to... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
Control Cloud IoT Gateway Sensors IoT Node Sensors IoT Node Cropland Area IoT Field Gateway IoT Actuator Fig. 2: Use cases of sensor measurement and actuator control A. Use Case Scenario The use case intends to explain precision agriculture from the perspective of the proposed IoT architecture [11]. Simply, sensors are... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
is below the standard range then it should start the irrigation process. In that case, the IoT ecosystem sends start instruction to the Salo actuator. Besides, the cropland sensors are sending live humidity measurements frequently to the cloud. At the same time, the IoT system continuously checks certain humidity level... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
role and integrated part of the precision agriculture system. With technological advancement (i.e., IoT, robotics, artificial intelligence, sensor technology, satellite imaging, GPS, image processing, etc.), precision agriculture allows better crops production by reducing human effort [12]. It enables farmers to monitor... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
Gateway. The usage of GPS in precision agriculture allows the analysis of soil property for croplands. And the analyzed results help to form field mapping and decide what type of soil is suitable for a given crop. The Mobile Clients linked to the IoT Field Gateway locally with the help of a Smart App and shortdistance r... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
the high-level description and points all of the components along with data flows. Here, we consider the functionalities of all gateways, actuators, and sensors as a process as they accept input data, perform some actions and produce some output information. In the DFD, we have used the circle to denote process, arrow t... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
(especially cyber security) in precision agriculture. Threat modeling is a proactive way to identify, enumerate, and prioritize threats thus helping to take appropriate safeguards against threats. Simply, it is formed to answer the questions like “Where are the potential threats to the system?”, “What are the most rele... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
placed in the croplands and trusted thus kept out of scope. Moreover, we believe that the Cloud/Satellite Imaging service providers take appropriate countermeasures for possible security issues as they are accounted for. Finally, we consider only threats that associated with major components, data flows (interaction), a... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
something Spoofing (Authentication) Modifying data or code on disk, network, memory, or elsewhere Tampering (Integrity) Claiming to have not performed an action Repudiation (Non-repudiation) Exposing information to someone not authorized to access it Information Disclosure (Confidentiality) Deny or degrade service to u... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
Actuator, Mobile Clients, MC to IFG, IFG to MC Denial of Service 14 Remote Users, RU to Cloud, Cloud to RU, IoT Gateway/Fog Node, Cloud to IGFN, IGFN to Cloud, IoT Field Gateway, IFG to IGFN, IGFN to IFG, Actuator, Actuator to IGFN, IGFN to Actuator, MC to IFG, IFG to MC Elevation of Privilege 04 IoT Gateway/Fog Node, ... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
TMM); Visual, Agile, and Simple Threat (VAST) Modeling. Among all, we have chosen the Microsoft STRIDE model to discover threats and possible mitigations because it is widely accepted in industry and academia. Moreover, an opensource (free of cost) tool called “Microsoft Threat Modeling Tool” is available from Microsof... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
access control, authentication, authorization, and more. For securing the data flow and at rest, several defense mechanisms are applied, for example, Hash-based Message Authentication Code (HMAC), Cipherbased MAC (CMAC), Rivest–Shamir–Adleman (RSA) cryptosystem, Advanced Encryption Standard (AES), Elliptic curve cryptog... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
the Microsoft STRIDE model to tackle cyber security issues. This work uses the Microsoft STRIDE model to identify, enumerate, and categorize threats for selected components, data flow, and external interactors. As an outcome, there identified fifty-eight cyber security threats which need to be controlled for a thriving sm... | {
"Authors": "Md. Rashid Al Asif, Khondokar Fida Hasan, Md Zahidul Islam, Rahamatullah Khondoker",
"Published": "2022-01-30",
"Summary": "The concept of traditional farming is changing rapidly with the introduction\nof smart technologies like the Internet of Things (IoT). Under the concept of\nsmart agriculture, ... |
A Game Theoretic Analysis for Cooperative Smart Farming Deepti Gupta∗§, Paras Bhatt†§, Smriti Bhatt‡ ∗Dept. of Computer Science, University of Texas at San Antonio, San Antonio, Texas 78249, USA †Dept. of Information Systems and Cyber Security, University of Texas at San Antonio, San Antonio, Texas 78249, USA ‡Dept. of... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
applications. However, complications may arise in CSF when some of the farms do not transfer high-quality data and rather rely on other farms to feed ML models. Another possibility is the presence of rogue farms in CSFs that want to snoop on other farms without actually contributing any data. In this paper, we analyze ... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
recent recognition of the World Food Program (WFP)1 as the Nobel Peace Prize recipient underlines the importance food security plays in our lives. Consequently, it is equally important to ensure that sustainable farming practices are followed so that land is not overused and exploited. Internet of Things (IoT) technolo... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
technologies as cloud and edge computing leads to the creation of smart farms which generates large amounts of data. This large amount of data can be analyzed using ML and Artificial Intelligence (AI) technologies to extract crucial insights and implications. The need for such smart farms has been stressed in research a... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
or better land use, or even the discovery of efficient seed application techniques. The use of cloud computing as well as web-based services has ushered in a new era where smart farms plan better yield management, adopt responsible farming practices, and indulge in sustainable techniques. It is for this reason that IT h... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
other. This type of cooperation between several smart farms is known as Cooperative Smart Farming (CSF). While CSF is an attractive framework to benefit multiple farms together, it is important to understand that there are also security and privacy issues that could arise in this framework. Data privacy could be a major... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
smart farms which generate low-quality data to build machine learning model can take advantage from other farms which have high-quality data. 2) We present two different use case scenarios including different types of smart farms in a CSF along with rationality assumption. 3) We propose a fair strategy to enforce farms... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
privacy issues in CSFs, and game theory application in the context of smart farming. A. Cooperative Smart Farming Farming has always been a community activity and in its truest sense a communal one that includes a participatory process. Therefore, even though individual farms operating alone may be efficient revenue max... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
to keep track of the covered land, monitor crop health and even classify weeds [11]. Moreover, the use of technology in smart farms has been so far advanced as distinguishing between different plowing techniques and classifying them according to plowing depth [12]. The use of Unmanned Aerial Vehicles (UAV) has been con... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
is a sizable generation of continuous data streams in smart and cooperative farms. Big data technologies play an important Communication and data exchange . . . . . . . . . . . Member Farm-1 Member Farm-2 Member Farm-3 Member Farm-N Physical Layer Edge Layer Cloud Layer BLE, Zigbee, Wi-Fi BLE, Zigbee, Wi-Fi MQTT, HTTP ... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
Another interesting application of AI capabilities in the smart farming context has been for weed classification where spectral images collected by small Micro Aerial Vehicles (MAVs) have been used to identify the weeds that may pose a danger to the surrounding plants or crops [19]. B. Security and Privacy issues on Coo... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
harvest or withering of plants because of inefficient application of fertilizers. Therefore, it is crucial to ensure security and privacy be robust in smart farms. Security and privacy issues in CSF architecture relate to authorization and trust, authentication and sever communication as well as compliance and regulatio... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
farms, another security challenge that arises is that these farms inadvertently play host to a plethora of security challenges [32]. However, enforcement of data transfer agreements and verifying the approved learning received from the cloud servers is a very important task in the connected farm system. Further, the is... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
be ported into a collaborative mechanism that can effectively serve participating smart farms in a cooperative bloc. Ensuring proper device authentication of smart devices with low computing power, that are embedded in connected farm systems, can also act as an effective deterrent against cyber-attacks and minimize sec... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
three sub-units. Gupta et al. [40] presents a novel fair strategy for collaborative deep learning game, where all mobile edge devices [41] choose any strategy to make own profit. The past studies show that no research has analyzed the farm’s rational behavior in CSF. Therefore, we develop a game model for farms in CSF a... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
smart farms in a CSF bloc which is based on a game theory-based model. Participating smart farms are rational choice makers and will only become members of CSF bloc if they get tangible benefits from it. These benefits have to be more beneficial than would normally accrue to the farms if they operated alone based on their... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
ML model is trained on the cloud or edge server with data that is transferred to it from the participating farms. Generally, we assume that the larger the dataset collected from different farms the better ML model will be; thus, there will be an inherent incentive for the participating farms to become a member of CSF b... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
being a member of a CSF bloc then, it can choose to defect from the bloc. However, we envision that these scenarios are only fringe options that depends on different types of participating farms and may or may not be seen often in the context of our smart farms based CSF architecture. B. Threat Model 1) Scenario 1: In ... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
good crop yield, predict scarcity of soil nutrients etc. The ML model shares the data with all the farms irrespective of their size or quality of the data contribution. In this scenario, Farm A has good devices, data collected from those devices, and resources; however, Farm B does not have good farm devices, data, and... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
4 depicts the second threat scenario where one of the farms could have malicious intents. Here, the scenario also consists of two farms, Farm A and Farm B, where Farm A is a technologically advanced farm compared to Farm B and Farm A provides better quality data to the ML model, similar to scenario 1. Now, Farm B decid... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
the smart farms receive models and their outputs/insights that accurately represent the aggregate learning from CSF bloc. C. Rationality Assumption Prior research in smart farming [46] have presented cyberattacks on smart farming infrastructure where farmers or IoT devices are controlled by adversary. Malicious partici... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
farms via AI-assisted application, where exists a social dilemma for all defection behavior. Table 1 shows game approach between two players, one has high-quality data and other has low-quality data. List of symbols are shown in Table 2. A. Game Theoretic Model Game theory allows for modeling situations of conflict and ... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
farms. IoT devices through edge gateways to build ML model in CSF game G. • Strategy (S): Each participant Pi can choose between two actions si (i) Cooperative (CP) or (ii) Defective (DF). Hence the set of strategies in this game is S = {CP,DF}. Strategy of each participant Pi determines whether Pi participates in CSF.... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
cm, cm′, storage cost cs, and membership cost co, and also pay penalty cp if any player breaks co-op membership agreement. Total cost ct i to participate in CSF can be characterized as ct i = co + cp + cm + cm′ + cs (1) Here, the benefit and the cost are not on the same scale as the first depends on the accuracy of ML mo... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
Total cost for participating in cooperative smart farming Ci Number of cooperative participants N −Ci Number of defective participants Farms are compared on a predefined set of attributes Soil Data Moisture Nutrient content Irrigation Data Water quality Watering schedule Yield Data Yield per hectare Yield history Shari... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
Nash Equilibrium in prisoners’ dilemma, is a mutual defection. Member farm-1 AWS IoT Greengrass AWS Lambda Raspberry Pi AWS IoT Gateway Devices/Sensors AWS IoT & other services Member farmers group-1 Communication and data exchange between users and applications BLE, Zigbee, Wi-Fi BLE, Zigbee, Wi-Fi MQTT, HTTP MQTT, HT... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
any results from cloud, i.e., no membership fees co, no communication between IoT devices and cloud. So, participants do not pay any communication costs cm, cm′, and storage cost cs. Now each participant Pi trains local data sets Di to build its ML model individually and pays only computation cost cplocal. None of part... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
by Equation (2). Hence, if a participant will deviate from the cooperation and play defection unilaterally, its payoff would be equal to Equation (3), which is always greater than cooperative payoff. Hence, each participant has incentive to deviate unilaterally and increases its payoff. Then, the All cooperate-CP strat... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
of sample data attributes is shown in Figure 5. This shows different categories of data items and hierarchy of attributes soil data, irrigation data, yield data, and sharing schedule for data and resources. Each of these categories are further broken down into other attributes that provide detailed information about th... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
of the game, each participant has to choose his strategy between CP and DG to play the CSF game G. However, in the beginning of this game, each participant is in dilemma to choose strategy, which can depend on other participant’s strategy. In CSF, each member farm does not know about the quality and quantity of other p... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
at co-op level. 6: if participant i belongs a cluster with at least one other participant j then 7: Pi , Pj ∈CP 8: else 9: Pi ∈DF 10: end if VI. PROPOSED IMPLEMENTATION In this section, we present a proposed implementation for CSF using proposed fair strategy utilizing AWS cloud and IoT platform, shown in Figure 6. In ... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
AWS IoT core provides mutual authentication based on certificates and encryption for secure data transfer and also enable different access control levels, which have been discussed in [27], [28]. The data can be stored at AWS cloud storage and be analyzed at the co-op level for classified member farms based on the qualit... | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
we plan to do further research on data attributes identification for ML models that can be used by CSFs to better classify the smart farms based on the quality and quantity of data. We also plan to implement the model with farming data set and evaluate the accuracy of ML model using proposed fair strategy. | {
"Authors": "Deepti Gupta, Paras Bhatt, Smriti Bhatt",
"Published": "2020-11-22",
"Summary": "The application of Internet of Things (IoT) and Machine Learning (ML) to the\nagricultural industry has enabled the development and creation of smart farms\nand precision agriculture. The growth in the number of smart f... |
An Efficient Data Warehouse for Crop Yield Prediction Vuong M. Ngo, Nhien-An Le-Khac, M-Tahar Kechadi School of Computer Science, College of Science, University College Dublin, Belfield, Dublin 4, Ireland Abstract. Nowadays, precision agriculture combined with modern information and communications technologies, is beco... | {
"Authors": "Vuong M. Ngo, Nhien-An Le-Khac, M-Tahar Kechadi",
"Published": "2018-06-26",
"Summary": "Nowadays, precision agriculture combined with modern information and\ncommunications technologies, is becoming more common in agricultural activities\nsuch as automated irrigation systems, precision planting, va... |
and very large. In particular, agricultural data is considered as Big Data in terms of volume, variety, velocity and veracity. Designing and developing a data warehouse for precision agriculture is a key foundation for establishing a crop intelligence platform, which will enable resource efficient agronomy decision mak... | {
"Authors": "Vuong M. Ngo, Nhien-An Le-Khac, M-Tahar Kechadi",
"Published": "2018-06-26",
"Summary": "Nowadays, precision agriculture combined with modern information and\ncommunications technologies, is becoming more common in agricultural activities\nsuch as automated irrigation systems, precision planting, va... |
forecasted to reach a total of 9.8 billion . Hence, The Pre-Print Paper was accepted in ICPA-14th June 24 – June 27, 2018 Montreal, Quebec, Canada Page 2 to satisfy the massively increased demand for food, crop yields must be significantly increased by using new farming technologies or methods. In agriculture, the key ... | {
"Authors": "Vuong M. Ngo, Nhien-An Le-Khac, M-Tahar Kechadi",
"Published": "2018-06-26",
"Summary": "Nowadays, precision agriculture combined with modern information and\ncommunications technologies, is becoming more common in agricultural activities\nsuch as automated irrigation systems, precision planting, va... |
final question, we usually need the crop price, the supply, and demand information within the last few years, etc. With technological advances in the area of information and communication technology (ICT), farmers can access and share valuable information and knowledge. They can also obtain knowledge on new research re... | {
"Authors": "Vuong M. Ngo, Nhien-An Le-Khac, M-Tahar Kechadi",
"Published": "2018-06-26",
"Summary": "Nowadays, precision agriculture combined with modern information and\ncommunications technologies, is becoming more common in agricultural activities\nsuch as automated irrigation systems, precision planting, va... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.