查看更多>>摘要：? 2022 Elsevier B.V.The implementation of artificial intelligence (AI) in parallel with remote sensing could be a powerful tool to manage irrigation scheduling on crops with narrow thresholds between water stress levels, such as cherry trees. This research assessed the water status of 'Regina' cherry trees using machine learning (ML) modeling from data extracted automatically using infrared thermal imagery (IRTI). These models were used to predict stomatal conductance (gs) and stem water potential (Ψs) (Model 1) and a complete assessment using a matrix differential analysis procedure per IRTI of cherry tree canopies' temperature and relative humidity (Model 2). Results showed that the supervised ML regression models presented high and significant correlation coefficients (R = 0.83 and R = 0.81, respectively) without signs of overfitting assessed through their performance. The complex interactions among climatic factors, the soil moisture, and canopy architecture observed in cherry trees or any other fruit tree oblige exploring the performance of ML-based models to offer simple alternatives for decision-making processes in the field.
查看更多>>摘要：? 2022 The AuthorsSustainability is a major challenge in agri-food systems. Digital technologies, such as Internet of Things (IoT) hold substantial promises for attaining the sustainability goals of the economy, environment and society at large. However, in practice it is difficult to evaluate to which extent these technologies contribute to sustainable development raising doubts about their impact. This paper demonstrates a stepwise approach that allows for measuring and monitoring IoT contribution to sustainability in a real-life context. The UN sustainable development goals (SDGs) underpin the principles of the approach by a typology and by framing the sustainability impact in terms of business opportunities. The approach has been developed and evaluated by 33 use cases in the EU-funded IoF2020 project. The research illustrates how the measurement and monitoring tool is applied in 5 of these use cases from different agricultural subsectors showing how the approach is applied and validated. The results indicate an overall positive impact of IoT on improving sustainability, although these results are also partly determined by other influential external factors that cannot be easily discerned in a practical situation. The main contribution of this approach is the set of instruments for practitioners to measure and monitor the impact of fast-changing technologies such as IoT to sustainability in a real-life context. This set of instruments can also be used by other stakeholders in large IoT projects where strategic sustainability objectives should be supported by IoT solutions. The stepwise approach is easy to communicate and supports stakeholders such as farmers in decision-making, but also policy makers and investors in funding projects.
查看更多>>摘要：? 2022 Elsevier B.V.With the rapid development of remote sensing (RS) technology, remote sensing images provide an important data basis for soil type mapping. In remote sensing images, temporal factor is difficult to obtain directly, and the rich geometric features are not used adequately. Multi-temporal remote sensing data could effectively reflect the temporal variation of ground objects, while how to extract multi-temporal image features more effectively for soil type interpretation needs to be studied. Moreover, it is not clear whether multi-temporal features and texture features can be effectively integrated to improve mapping accuracies. Therefore, taking five soil types of Laoshan County, Shandong Peninsula, China as the subject investigated and six remote sensing images as data sources, this paper explored and compared two extraction methods of multi-temporal features from remote sensing images. The effects of the eight different texture features fused the multi-temporal features on digital soil mapping were also analyzed. The results showed that the principal component extraction result based on the tasseled cap transformation was better than based on the spectral band synthesis, increasing the overall accuracy by 3.83–11.41% and the kappa index by 0.03–0.13. The fusion of multi-temporal features and texture features can effectively improve accuracies of soil type mapping. After the addition of correlation texture feature parameter, the overall accuracy (86.81%) and Kappa index (0.81) were increased by 11.92% and 0.16, respectively. The research results showed that multi-temporal features in remote sensing images had great advantages in digital soil mapping, and the effective fusion with texture features provided a new idea for improving the accuracy of digital soil mapping.
查看更多>>摘要：? 2022 Elsevier B.V.Fruit detection and localization is of great significance for horticulture work and robotic harvesting in orchards. Although the existing studies of fruit detection have achieved good results based on 2D image analysis, accurate fruit detection on trees is still challenging because of illumination changes, shielding of leaves and branches, overlapping of fruits and so on. To improve the accuracy of fruit detection and location, this paper proposes a novel ripe pomegranate fruit detection and location method based on improved F-PointNet and 3D clustering method, which is consisting of: (1) RGB-D feature fusion Mask R-CNN was used to realize fruit detection and segmentation; (2) PointNet combined with OPTICS algorithm based on manifold distance and PointFusion was used to segment point clouds in the frustum fruit region, and 3D box was placed in the region of interest; (3) The sphere fitting was performed to obtain the position and the size of a pomegranate. The comparative experiments have been carried out and analyzed, the RGB-D feature fusion Mask R-CNN has the best performance with the F1 score of 0.845 and the AP score of 0.952 respectively, and the improved F-PointNet has better performance than the classical F-PointNet. The measurement radius experiment results of 100 pomegranate samples randomly selected demonstrate that the RMSE is 0.235 cm, the R2 is 0.826, and the position error is less than 5 mm. These results validate that the proposed detection and location method can effectively detect and locate a single ripe pomegranate under unstructured orchard environment.
查看更多>>摘要：? 2022 The Author(s)Honeybees plays vital role for the environmental sustainability and overall agricultural economy. Assisting bee colonies within their proper functioning brings the attention of researchers around the world. Electronics systems and machine learning algorithms are being developed for classifying specific undesirable bee behaviors in order to alert about upcoming substantial losses. However, classifiers could be impaired when used for general honeybee colony state inference. Application of the classifier models for the hazardous situations detection without focusing on the model's genericity could result with systems that are not applicable in the real environment. Furthermore, the detection of a specific phenomenon does not provide researchers with any new conclusions about the honeybee colony life but only with the binary information about hazardous situation presence. In our research we propose a method for inferring the bee colony state using a sensitive contrastive autoencoder and an anomaly detection model. With presented approach, hive's internal state is modeled with the use of an autoencoder's latent vector extended with in-hive temperature dynamics. We test our methodology with a bee feeding experiment where the glucose syrup application was detected and the length of food intake was estimated. As our methodology has been applied successfully, we argue that contrastive autoencoders can be used for precise inference about the behavior of honeybees.
查看更多>>摘要：? 2022The advent of digital technologies has brought substantial improvements in various domains. This article provides a comprehensive review of research emphasizing AI-enabled IoT applications in poultry health and welfare management. This study focused on poultry welfare since modern poultry management is confronted with issues relating to standardized parameters for welfare assessment and robust monitoring systems, particularly for broilers' health and disease outbreak prevention. Evidence has shown that modern digital technologies have high possibilities for intelligent automation of current and future poultry management operations to facilitate high-quality and low-cost poultry production. Therefore, this study presents a systematic review of the current state-of-the-art AI-enabled IoT systems and their recent advances in developing intelligent systems in this domain. Also, the study provides an overview of the critical applications of identified digital technologies in poultry welfare management. Lastly, the study discusses the challenges and opportunities of AI and IoT in poultry farming.
查看更多>>摘要：? 2022Grape berry ripening is a complex process, and predicting the quality of wine starting from the ripening kinetics of grape berries is a challenging task. To tackle this problem, we present a decision-support system based on coupling expert know-how with probability laws encapsulated in a probabilistic model, a dynamic Bayesian network. The proposed approach predicts the ripening kinetics of grape berries starting from initial measurements and weather conditions, and then exploits the information to evaluate the potential of the wine that will produced from them. The results show that the dynamic Bayesian network predicts the total acidity concentration and the sugar content of the grape berries with a small amount of error (mean of 6% for total acidity concentration, 10% for sugar content) that is considered satisfying by the experts, making it possible to predict the ideal moment for harvesting the grapes up to two weeks in advance. Moreover, feeding the results from the probabilistic model to a fuzzy expert model, the predicted trajectories are compared to an ideal trajectory described by wine experts and formalized mathematically. From this comparison, it is possible to anticipate drifts in wine sensory quality right from the step of grape ripening.
查看更多>>摘要：? 2022 Elsevier B.V.Reinforcement learning (RL), including multi-armed bandits, is a branch of machine learning that deals with the problem of sequential decision-making in uncertain and unknown environments through learning by practice. While best known for being the core of the artificial intelligence (AI) world's best Go game player, RL has a vast range of potential applications. RL may help to address some of the criticisms leveled against crop management decision support systems (DSS): it is an interactive, geared towards action, contextual tool to evaluate series of crop operations faced with uncertainties. A review of RL use for crop management DSS reveals a limited number of contributions. We profile key prospects for a human-centered, real-world, interactive RL-based system to face tomorrow's agricultural decisions, and theoretical and ongoing practical challenges that may explain its current low uptake. We argue that a joint research effort from the RL and agronomy communities is necessary to explore RL's full potential.
查看更多>>摘要：? 2022 Elsevier B.V.In agricultural image analysis, optimal model performance is keenly pursued for better fulfilling visual recognition tasks (e.g., image classification, segmentation, object detection and localization), in the presence of challenges with biological variability and unstructured environments. Large-scale, balanced and ground-truthed image datasets are tremendously beneficial but most often difficult to obtain to fuel the development of highly performant models. As artificial intelligence through deep learning is impacting analysis and modeling of agricultural images, image augmentation plays a crucial role in boosting model performance while reducing manual efforts for image collection and labelling, by algorithmically creating and expanding datasets. Beyond traditional data augmentation techniques, generative adversarial network (GAN) invented in 2014 in the computer vision community, provides a suite of novel approaches that can learn good data representations and generate highly realistic samples. Since 2017, there has been a growth of research into GANs for image augmentation or synthesis in agriculture for improved model performance. This paper presents an overview of the evolution of GAN architectures followed by a first systematic review of various applications in agriculture and food systems (https://github.com/Derekabc/GANs-Agriculture), involving a diversity of visual recognition tasks for plant health conditions, weeds, fruits (preharvest), aquaculture, animal farming, plant phenotyping as well as postharvest detection of fruit defects. Challenges and opportunities of GANs are discussed for future research.
查看更多>>摘要：? 2022 Elsevier B.V.The slope (g1) of stomatal conductance to photosynthesis is an important parameter in the optimal stomatal behavior theory-based stomatal conductance model of Medlyn et al. (2011). Although studies have modelled the spatial variations in g1, disclosing its variations over environmental gradients and different plant functional types. However, the above methods are still not accurate enough on a global scale, as they do not consider the temporal variations in g1. To address this issue we used the Ensemble Kalman Filter (EnKF) to assimilate tower-based gross primary productivity (GPP) and latent heat flux (LE) of 17 cropland flux sites into a remote sensing (RS)-based evapotranspiration-photosynthesis coupled model, termed SCOPES-Crop, to derive the temporal variations in g1 for C3 and C4 crops. We also used the feedforward artificial neural network (FANN) along with RS variables to model g1. Results showed g1 to rise rapidly in spring and summer, and then decline in autumn. The value of g1 reached the lowest value and remained stable in wintertime. FANN-based modeling of g1 showed R(RMSE) = 0.81 (1.94 kPa0.5) and 0.90 (0.70 kPa0.5) for C3 and C4 Crops, respectively, for the testing dataset. The estimates of GPP and LE using FANN-derived g1 at the 17 flux sites were improved as compared to that using fixed g1. The mean values of site-level R(RMSE) for GPP and LE simulated using FANN-derived g1 are 0.92 (1.8 gC m?2 d-2) and 0.85 (22.5 W m?2), respectively. Our results revealed notable seasonal variations in g1, indicating the importance of considering the temporal variations in g1 in evapotranspiration-photosynthesis coupled model. The FANN along with RS variables showed great potential of representing the g1 variations.
NSTL
Elsevier