An Empirical Multi-classifier for Coffee Rust Detection in Colombian Crops

2020

rev.ing.univ.Medellin

Early warning systems are designed to inform the largest number of users, such as a country or a region, about a risky situation. However, in specific domains such as agriculture, it is commonly required for these alerts to be more specific according to the crops location and their properties. Consequently, the web services of these systems must be adapted. On the other hand, the enterprise services bus, with its mediation capabilities (such as message transformation and routing), and complex event processing with their monitoring characteristics, can be integrated to meet the adaptation requirements of web services at runtime. This paper presents an improvement for an early warning system for coffee production that, according to the area in which a crop is located and its phenology, manages the adaptation of alerts for coffee rust, based on the integration of an enterprise services bus and a complex events processing.

Section: Application Domainmentioning

confidence: 99%

Web Service Adaptation for the Customization of Early Alerts in Agriculture

Aguilar

Lasso

2020

rev.ing.univ.Medellin

“…A real case is presented in [80,82]. Their dataset includes 147 instances to try to detect the incidence rate of rust.…”

Section: Amount Of Datamentioning

confidence: 99%

“…On the basis of the foregoing, we have identified four data quality issues (noise, incompleteness, outliers and amount of data) in a real dataset for coffee rust detection exposed in [80,82]. The data used in this work was collected at the Technical Farm (Naranjos) of the Supracafe, in Cajibio, Cauca, Colombia (21°35'08"N, 76°32'53"W), during 2011-2013.…”

Section: Timelinessmentioning

confidence: 99%

A systematic review of data quality issues in knowledge discovery tasks

Ledezma

Rev. ing. univ. Medellín

2016

Self Cite

Large volume of data is growing because organizations are continuously capturing the collective amount of data for a better decision-making process. The most fundamental challenge is to explore the large volumes of data and extract useful knowledge for future actions through knowledge discovery tasks, nevertheless many data has poor quality. We presented a systematic review of the data quality issues in knowledge discovery tasks and a case study applied to agricultural disease named coffee rust. ResumenHay un gran crecimiento en el volumen de datos porque las organizaciones capturan permanentemente la cantidad colectiva de datos para lograr un mejor proceso de toma de decisiones. El desafío mas fundamental es la exploración de los grandes volúmenes de datos y la extracción de conocimiento útil para futuras acciones por medio de tareas para el descubrimiento del conocimiento; sin embargo, muchos datos presentan mala calidad. Presentamos una revisión sistemática de los asuntos de calidad de datos en las áreas del descubrimiento de conocimiento y un estudio de caso aplicado a la enfermedad agrícola conocida como la roya del café.

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

“…Model stacking is an ensemble method that allows one to improve the model predictions by combining the output of different individual models [55]. While model stacking has been used in various applications linked to agricultural monitoring using remote sensing data (e.g., [56]- [58]), to our knowledge, it has received very little attention for the estimation of crop cover parameters in heterogeneous smallholder agricultural landscapes. However, we can mention the work of [59], [60] and [61] on using a stacking approach for crop type mapping in South Africa, Zimbabwe and Mali, respectively.…”

Section: Introductionmentioning

confidence: 99%

Combining UAV and Sentinel-2 Imagery for Estimating Millet FCover in a Heterogeneous Agricultural Landscape of Senegal

Diack,

Diene,

Louise

et al. 2024

In recent decades, remote sensing has been shown to be useful for crop cover monitoring over smallholder agricultural landscapes such as agroforestry parklands. However, the fraction of green vegetation cover (FCover) has received little attention. Indeed, the collection of FCover ground data representative of the within-field heterogeneity is time-consuming. Thus, this study aims to bridge this gap by proposing an original methodological framework combining FCover data derived from unmanned aerial vehicle (UAV) and Sentinel-2 (S2) images for estimating millet FCover at the landscape scale in an agroforestry parkland of the groundnut basin of Senegal during the 2021 and 2022 cropping seasons. UAV-based FCover was computed over a 3 m x 3 m grid using a thresholding approach for 6 dates over the cropping seasons and then used as ground observation for the upscaling of millet FCover at the landscape scale with S2 data. Various spectral vegetation indices (VIs) and textural features (TFs) were derived from S2, and several modeling approaches based on machine learning algorithms were benchmarked. Our results showed that the modeling approach using the full time series in combination with a random forest algorithm was able to explain 73% (RMSE = 12.13%) of the UAV-FCover variability after validation in 2021 and 2022. Additionally, UAV images are suitable for consistent monitoring of millet FCover over heterogeneous agricultural landscapes by training S2 satellite images. To further check its robustness, this approach should be tested for different crops and practices across a variety of agricultural landscapes in Sub-Saharan Africa.