Canopy height and biomass prediction in Mombaça guinea grass pastures using satellite imagery and machine learning

Bretas, Igor L.; Valente, Domingos Sárvio Magalhães; Oliveira, Thiago Furtado de; Montagner, Denise Baptaglin; Euclides, Valéria Pacheco Batista; Chizzotti, M. L.

doi:10.1007/s11119-023-10013-z

Cited by 5 publications

(7 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The poor performance of the models in predicting dry AGB in tropical pastures has also been observed in previous studies, which reported R 2 values less than 0.30 17,19,32 . The low predictive ability of dry AGB in previous studies 17,19 was attributed to the low variability in the dry biomass dataset used for modeling, whose coe cient of variation was approximately 26% 19 . In the current study, the coe cient of variation for the observed dry AGB dataset was approximately 19% (Table 4), which could be a plausible explanation.…”

Section: Discussionsupporting

confidence: 79%

“…This grazing management allowed for a gap between eld collection and image availability of ± 10 days, which allowed adequate data collection free from cloud cover, which is the main limitation of satellite optical sensors. However, Bretas et al 19 observed that the predictive performance of the models was enhanced when the maximum interval between image acquisition and eld observation was restricted to one day instead of ve days. This information gap is signi cant in rotational grazing, where the impact of changing pasture conditions occurs in the short term during the growing season.…”

Section: Discussionmentioning

confidence: 99%

“…The exponential evolution of digital computers harnessed machine learning algorithms, which have been reported to frequently enhance predictive performance compared with simpler linear regression models 17,18 . Nonetheless, in tropical pastures, the use of satellites to estimate dry AGB has resulted in poor predictive performance 17,19 , which has been attributed to the presence of a high fraction of senescent material in the biomass and soil background scattering effects 20,21 . Therefore, the aboveground dry biomass of tropical pastures still needs to be addressed and investigated to build feasible models to implement in eld conditions.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Using Sentinel-2 satellite images and machine learning algorithms to predict tropical pasture forage biomass, crude protein, and fiber content

Fernandes,

Junior,

Adams

et al. 2023

Preprint

View full text Add to dashboard Cite

Grasslands cover approximately 24% of the Earth’s surface and are the main feed source for cattle and other ruminants. Sustainable and efficient grazing systems require regular monitoring of the quantity and quality of pastures. This study demonstrates the potential of estimating pasture leaf biomass, crude protein (CP) and fiber content of tropical pastures using Sentinel-2 satellite images and machine learning algorithms. Field datasets and satellite images were assessed from an experimental area of Marandu palisade grass pastures, with or without nitrogen fertilization, and managed under continuous stocking during the pasture growing season from 2016 to 2020. Models based on support vector regression (SVR) and random forest (RF) machine-learning algorithms were developed using meteorological data, spectral reflectance, and vegetation indices (VI) as input features. In general, SVR slightly outperformed the RF models. The best predictive models to estimate biomass were those with VI combined with meteorological data. For CP and fiber content, the best predictions were achieved using a combination of spectral bands and meteorological data, resulting in R2 of 0.66 and 0.57, and RMSPE of 0.03 and 0.04 g/g dry matter. Our results have promising potential to improve precision feeding technologies and decision support tools for efficient grazing management.

show abstract

Section: Discussionsupporting

confidence: 79%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Using Sentinel-2 satellite images and machine learning algorithms to predict tropical pasture forage biomass, crude protein, and fiber content

Fernandes,

Junior,

Adams

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The global accuracy represents the percentage of correct predictions of the model. However, relying solely on accuracy is insufficient for a complete evaluation of classification tasks with unbalanced classes (Bretas et al, 2023). Sensitivity refers to the model's ability to detect true positives, while specificity refers to the model's ability to detect true negatives (Hastie et al, 2009;Kamilaris & Prenafeta-Boldú, 2018).…”

Section: Model Evaluationmentioning

confidence: 99%

Detection and mapping of Amaranthus spinosus L. in bermudagrass pastures using drone imagery and deep learning for a site‐specific weed management

Bretas,

Dubeux,

Zhao

et al. 2024

Agronomy Journal

View full text Add to dashboard Cite

Weed encroachment negatively affects pasture productivity by reducing herbage allowance, stocking rates, and livestock performance. Amaranthus spinosus L. is a weed species widely found in pastures worldwide and is considered challenging for ranchers due to its great potential for invasion, making it difficult to control. The high costs of chemical application and the global concern about environmental impacts restrict indiscriminate herbicide spraying in pastures. Site‐specific weed management (SSWM) is a weed management strategy based on weed spot‐spraying that has the potential to overcome these issues. Images from unmanned aerial vehicles (UAVs) can provide valuable information for weed mapping to drive the herbicide application in pastures. Deep learning techniques have been highlighted in image classification tasks. We developed a deep convolutional neural network (CNN)‐based image segmentation model based on the U‐Net architecture to detect and map Amaranthus spinosus in bermudagrass pastures using red–green–blue images acquired through UAV flying in moderate‐high altitude. The images were acquired from twelve paddocks under three treatments (weed‐free, weed‐strips, or weed‐infested) during the summer (2021‐2022). The CNN model was able to detect around 80% of the A. spinosus with an average prediction accuracy of 94%. Our weed mapping showed the potential of using the U‐Net model to generate a herbicide application map to be inserted into the sprayer system, reducing up to 76% of the amount of herbicide applied. Further studies are encouraged to increase the robustness of the model across species and development stages and develop sprayer systems to implement the spot‐spraying in field conditions.

show abstract

“…The exponential evolution of digital computers harnessed machine learning algorithms, which have been reported to frequently enhance predictive performance compared with simpler linear regression models 17 , 18 . Nonetheless, in tropical pastures, the use of satellites to estimate FM has resulted in poor predictive performance 17 , 19 , which has been attributed to the presence of a high fraction of senescent material in the biomass and soil background scattering effects 20 , 21 . Therefore, the dry FM of tropical pastures still needs to be addressed and investigated to build feasible models to implement in field conditions.…”

Section: Introductionmentioning

confidence: 99%

Using sentinel-2 satellite images and machine learning algorithms to predict tropical pasture forage mass, crude protein, and fiber content

Fernandes,

FernandesJunior,

Adams

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Grasslands cover approximately 24% of the Earth’s surface and are the main feed source for cattle and other ruminants. Sustainable and efficient grazing systems require regular monitoring of the quantity and nutritive value of pastures. This study demonstrates the potential of estimating pasture leaf forage mass (FM), crude protein (CP) and fiber content of tropical pastures using Sentinel-2 satellite images and machine learning algorithms. Field datasets and satellite images were assessed from an experimental area of Marandu palisade grass (Urochloa brizantha sny. Brachiaria brizantha) pastures, with or without nitrogen fertilization, and managed under continuous stocking during the pasture growing season from 2016 to 2020. Models based on support vector regression (SVR) and random forest (RF) machine-learning algorithms were developed using meteorological data, spectral reflectance, and vegetation indices (VI) as input features. In general, SVR slightly outperformed the RF models. The best predictive models to estimate FM were those with VI combined with meteorological data. For CP and fiber content, the best predictions were achieved using a combination of spectral bands and meteorological data, resulting in R2 of 0.66 and 0.57, and RMSPE of 0.03 and 0.04 g/g dry matter. Our results have promising potential to improve precision feeding technologies and decision support tools for efficient grazing management.

show abstract

Canopy height and biomass prediction in Mombaça guinea grass pastures using satellite imagery and machine learning

Cited by 5 publications

References 48 publications

Using Sentinel-2 satellite images and machine learning algorithms to predict tropical pasture forage biomass, crude protein, and fiber content

Using Sentinel-2 satellite images and machine learning algorithms to predict tropical pasture forage biomass, crude protein, and fiber content

Detection and mapping of Amaranthus spinosus L. in bermudagrass pastures using drone imagery and deep learning for a site‐specific weed management

Using sentinel-2 satellite images and machine learning algorithms to predict tropical pasture forage mass, crude protein, and fiber content

Contact Info

Product

Resources

About