Agriculture is essential for global income, poverty reduction, and food security, with crop yield being a crucial measure in this field. Traditional crop yield prediction methods, reliant on subjective assessments such as farmers’ experiences, tend to be error-prone and lack precision across vast farming areas, especially in data-scarce regions. Recent advancements in data collection, notably through high-resolution sensors and the use of deep learning (DL), have significantly increased the accuracy and breadth of agricultural data, providing better support for policymakers and administrators. In our study, we conduct a systematic literature review to explore the application of DL in crop yield forecasting, underscoring its growing significance in enhancing yield predictions. Our approach enabled us to identify 92 relevant studies across four major scientific databases: the Directory of Open Access Journals (DOAJ), the Institute of Electrical and Electronics Engineers (IEEE), the Multidisciplinary Digital Publishing Institute (MDPI), and ScienceDirect. These studies, all empirical research published in the last eight years, met stringent selection criteria, including empirical validity, methodological clarity, and a minimum quality score, ensuring their rigorous research standards and relevance. Our in-depth analysis of these papers aimed to synthesize insights on the crops studied, DL models utilized, key input data types, and the specific challenges and prerequisites for accurate DL-based yield forecasting. Our findings reveal that convolutional neural networks and Long Short-Term Memory are the dominant deep learning architectures in crop yield prediction, with a focus on cereals like wheat (Triticum aestivum) and corn (Zea mays). Many studies leverage satellite imagery, but there is a growing trend towards using Unmanned Aerial Vehicles (UAVs) for data collection. Our review synthesizes global research, suggests future directions, and highlights key studies, acknowledging that results may vary across different databases and emphasizing the need for continual updates due to the evolving nature of the field.

This paper provides an in-depth literature review on image processing techniques, focusing on deep learning approaches for anomaly detection and classification in photovoltaics. It examines key components of UAV-based PV inspection, including data acquisition protocols, panel segmentation and geolocation, anomaly classification, and optimizations for model generalization. Furthermore, challenges related to domain adaptation, dataset limitations, and multimodal fusion of RGB and thermal data are also discussed. Finally, research gaps and opportunities are analyzed to create a holistic, scalable, and real-time inspection workflow for large-scale installation. This review serves as a reference for researchers and industry professionals to advance UAV-based PV inspection.

The Mediterranean forests, particularly Cork oak (Quercus suber L., 1927), make a major contribution to the fight against climate change through Carbon sequestration. Hence, there is a great interest in the accurate quantification of biomass and carbon stock. In this context, this study aims at assessing the performance of a new approach, based on the combination of Unmanned aerial vehicle airborne Aerial laser scanning (ALS-UAV) and Terrestrial laser scanning (TLS) data, in the determination of dendrometric parameters (Circumference at 1.30 m and Tree Height), and consequently the estimation of biomass and carbon stock, considering field data as reference. This study takes the Maamora forest in Morocco as an example of a Mediterranean Cork oak forest. The methodology consists of collecting data at three levels: the entire area level for an ALS-UAV scan, the plot and tree levels for TLS surveys, as well as field data collection. Afterwards, dendrometric parameters (Circumference at 1.30 m and the Tree height) were estimated using individual tree segmentation and biomass; the carbon stock (aboveground, belowground, and total) was estimated using allometric equations. The comparison of the estimated dendrometric parameters with those measured in the field shows a strong relationship, with a Pearson coefficient of 0.86 and 0.83, a correlation coefficient (R²) of 0.81 and 0.71, and a Root mean square error (RMSE) of 1.84 cm and 0.47 m, respectively. Concerning the biomass and carbon stock estimation, the proposed approach gives a satisfactory accuracy, with a Pearson coefficient of 0.77, an R2 of 0.83, and an RMSE of 36.40 kg for biomass and 20.24 kg for carbon stock.

Wetlands, among the most productive ecosystems on Earth, shelter a diversity of species and help maintain ecological balance. However, they are witnessing growing anthropogenic and climatic threats, which underscores the need for regular and long-term monitoring. This study presents a systematic review of 121 peer-reviewed articles published between January 2015 and 30 April 2025 that applied machine learning (ML) and deep learning (DL) for wetland mapping and bird-habitat monitoring. Despite rising interest, applications remain fragmented, especially for avian habitats; only 39 studies considered birds, and fewer explicitly framed wetlands as bird habitats. Following PRISMA 2020 and the SPIDER framework, we compare data sources, classification methods, validation practices, geographic focus, and wetland types. ML is predominant overall, with random forest the most common baseline, while DL (e.g., U-Net and Transformer variants) is underused relative to its broader land cover adoption. Where reported, DL shows a modest but consistent accuracy over ML for complex wetland mapping; this accuracy improves when fusing synthetic aperture radar (SAR) and optical data. Validation still relies mainly on overall accuracy (OA) and Kappa coefficient ($\kappa$), with limited class-wise metrics. Salt marshes and mangroves dominate thematically, and China geographically, whereas peatlands, urban marshes, tundra, and many regions (e.g., Africa and South America) remain underrepresented. Multi-source fusion is beneficial yet not routine; The combination of unmanned aerial vehicles (UAVs) and DL is promising for fine-scale avian micro-habitats but constrained by disturbance and labeling costs. We then conclude with actionable recommendations to enable more robust and scalable monitoring. This review can be considered as the first comparative synthesis of ML/DL methods applied to wetland mapping and bird-habitat monitoring, and highlights the need for more diverse, transferable, and ecologically/socially integrated AI applications in wetland and bird-habitat monitoring.

Climate change, population growth, and economic shocks govern a context where food security and economic sustainability represent major challenges for the agricultural sector. Research for innovative production systems that ensure a better allocation of resources is a necessity to provide the foundations for farm reconversion. In this way, we carried out our work relating to precision farming, which is one of the innovative approaches aimed at ensuring the sustainability of agricultural production systems, thanks to its application principles and potential benefits. This synthesis paper examines aspects of assessing the impact of the use of such technology by analyzing previous research. The analysis carried out showed that the study of the impact of the use of precision technologies focused on three essential components on a micro-economic scale: the economic component, the environmental component, and the agronomic component. Prior studies examining the advantages of precision technologies have mostly relied on the examination of experiments and the application of quantitative analysis methods to measure the impact on environmental, economic, and agronomic parameters. The results of the study demonstrated that the adoption of precision farming technologies has provided advantages that contribute to the sustainability of agricultural production systems. Specifically, reducing environmental impact, cutting GHG (greenhouse gases) emissions by over 80%, valorizing natural resources (water and soil) with irrigation water savings of over 26%, and improving production efficiency and effectiveness. However, we suggest further studies examining the effects of precision agriculture using an integrated approach to assess the agronomic, economic, environmental, and social aspects of a production system as a whole. These studies will provide recommendations for adapting precision agriculture technologies to a wide range of farm types. In turn, highlighting the benefits of using precision farming technologies will support the process of adoption by farmers. The overview and findings presented in this article should point researchers in the direction of further research into precision farming technologies and provide extension staff, farm advisors, and farm machinery dealers with guidelines for promoting the adoption of precision farming.

Spatialization of biomass and carbon stocks is essential for a good understanding of the forest stand and its characteristics, especially in degraded Mediterranean cork oak forests. Furthermore, the analysis of biomass and carbon stock changes and dynamics is essential for understanding the carbon cycle, in particular carbon emissions and stocks, in order to make projections, especially in the context of climate change. In this research, we use a multidimensional framework integrating forest survey data, LiDAR UAV data, and extracted vegetation indices from Landsat imagery (NDVI, ARVI, CIG, etc.) to model and spatialize cork oak biomass and carbon stocks on a large scale. For this purpose, we explore the use of univariate and multivariate regression modeling and examine several types of regression, namely, multiple linear regression, stepwise linear regression, random forest regression, simple linear regression, logarithmic regression, and quadratic and cubic regression. The results show that for multivariate regression, stepwise regression gives good results, with R² equal to 80% and 65% and RMSE equal to 2.59 and 1.52 Mg/ha for biomass and carbon stock, respectively. Random forest regression, chosen as the ML algorithm, gives acceptable results, explaining 80% and 60% of the variation in biomass and carbon stock, respectively, and an RMSE of 2.74 and 1.72 Mg/ha for biomass and carbon stock, respectively. For the univariate regression, the simple linear regression is chosen because it gives satisfactory results, close to those of the quadratic and cubic regressions, but with a simpler equation. The vegetation index chosen is ARVI, which shows good performance indices, close to those of the NDVI and CIG. The assessment of biomass and carbon stock changes in the study area over 35 years (1985–2020) showed a slight increase of less than 10 Mg/ha and a decrease in biomass and carbon stock over a large area.