Artificial Intelligence in Agriculture最新文献

Accurate watermelon yield estimation is crucial to the agricultural value chain, as it guides the allocation of agricultural resources as well as facilitates inventory and logistics planning. The conventional method of watermelon yield estimation relies heavily on manual labor, which is both time-consuming and labor-intensive. To address this, this work proposes an algorithmic pipeline that utilizes unmanned aerial vehicle (UAV) videos for detection and counting of watermelons. This pipeline uses You Only Look Once version 8 s (YOLOv8s) with panorama stitching and overlap partitioning, which facilitates the overall number estimation of watermelons in field. The watermelon detection model, based on YOLOv8s and obtained using transfer learning, achieved a detection accuracy of 99.20 %, demonstrating its potential for application in yield estimation. The panorama stitching and overlap partitioning based detection and counting method uses panoramic images as input and effectively mitigates the duplications compared with the video tracking based detection and counting method. The counting accuracy reached over 96.61 %, proving a promising application for yield estimation. The high accuracy demonstrates the feasibility of applying this method for overall yield estimation in large watermelon fields.

Knowledge of the factors influencing nutrient-limited subtropical maize yield and subsequent prediction is crucial for effective nutrient management, maximizing profitability, ensuring food security, and promoting environmental sustainability. We analyzed data from nutrient omission plot trials (NOPTs) conducted in 324 farmers' fields across ten agroecological zones (AEZs) in the Eastern Indo-Gangetic Plains (EIGP) of Bangladesh to explain maize yield variability and identify variables controlling nutrient-limited yields. An additive main effect and multiplicative interaction (AMMI) model was used to explain maize yield variability with nutrient addition. Interpretable machine learning (ML) algorithms in automatic machine learning (AutoML) frameworks were subsequently used to predict attainable yield relative nutrient-limited yield (RY) and to rank variables that control RY. The stack-ensemble model was identified as the best-performing model for predicting RYs of N, P, and Zn. In contrast, deep learning outperformed all base learners for predicting RY_K. The best model's square errors (RMSEs) were 0.122, 0.105, 0.123, and 0.104 for RY_N, RY_P, RY_K, and RY_Zn, respectively. The permutation-based feature importance technique identified soil pH as the most critical variable controlling RY_N and RY_P. The RY_K showed lower in the eastern longitudinal direction. Soil N and Zn were associated with RY_Zn. The predicted median RY of N, P, K, and Zn, representing average soil fertility, was 0.51, 0.84, 0.87, and 0.97, accounting for 44, 54, 54, and 48% upland dry season crop area of Bangladesh, respectively. Efforts are needed to update databases cataloging variability in land type inundation classes, soil characteristics, and INS and combine them with farmers' crop management information to develop more precise nutrient guidelines for maize in the EIGP.

Instance segmentation, an important image processing operation for automation in agriculture, is used to precisely delineate individual objects of interest within images, which provides foundational information for various automated or robotic tasks such as selective harvesting and precision pruning. This study compares the one-stage YOLOv8 and the two-stage Mask R-CNN machine learning models for instance segmentation under varying orchard conditions across two datasets. Dataset 1, collected in dormant season, includes images of dormant apple trees, which were used to train multi-object segmentation models delineating tree branches and trunks. Dataset 2, collected in the early growing season, includes images of apple tree canopies with green foliage and immature (green) apples (also called fruitlet), which were used to train single-object segmentation models delineating only immature green apples. The results showed that YOLOv8 performed better than Mask R-CNN, achieving good precision and near-perfect recall across both datasets at a confidence threshold of 0.5. Specifically, for Dataset 1, YOLOv8 achieved a precision of 0.90 and a recall of 0.95 for all classes. In comparison, Mask R-CNN demonstrated a precision of 0.81 and a recall of 0.81 for the same dataset. With Dataset 2, YOLOv8 achieved a precision of 0.93 and a recall of 0.97. Mask R-CNN, in this single-class scenario, achieved a precision of 0.85 and a recall of 0.88. Additionally, the inference times for YOLOv8 were 10.9 ms for multi-class segmentation (Dataset 1) and 7.8 ms for single-class segmentation (Dataset 2), compared to 15.6 ms and 12.8 ms achieved by Mask R-CNN's, respectively. These findings show YOLOv8's superior accuracy and efficiency in machine learning applications compared to two-stage models, specifically Mask-R-CNN, which suggests its suitability in developing smart and automated orchard operations, particularly when real-time applications are necessary in such cases as robotic harvesting and robotic immature green fruit thinning.

Machine learning and deep learning are subsets of Artificial Intelligence that have revolutionized object detection and classification in images or videos. This technology plays a crucial role in facilitating the transition from conventional to precision agriculture, particularly in the context of weed control. Precision agriculture, which previously relied on manual efforts, has now embraced the use of smart devices for more efficient weed detection. However, several challenges are associated with weed detection, including the visual similarity between weed and crop, occlusion and lighting effects, as well as the need for early-stage weed control. Therefore, this study aimed to provide a comprehensive review of the application of both traditional machine learning and deep learning, as well as the combination of the two methods, for weed detection across different crop fields. The results of this review show the advantages and disadvantages of using machine learning and deep learning. Generally, deep learning produced superior accuracy compared to machine learning under various conditions. Machine learning required the selection of the right combination of features to achieve high accuracy in classifying weed and crop, particularly under conditions consisting of lighting and early growth effects. Moreover, a precise segmentation stage would be required in cases of occlusion. Machine learning had the advantage of achieving real-time processing by producing smaller models than deep learning, thereby eliminating the need for additional GPUs. However, the development of GPU technology is currently rapid, so researchers are more often using deep learning for more accurate weed identification.

The transformation of age-old farming practices through the integration of digitization and automation has sparked a revolution in agriculture that is driven by cutting-edge computer vision and artificial intelligence (AI) technologies. This transformation not only promises increased productivity and economic growth, but also has the potential to address important global issues such as food security and sustainability. This survey paper aims to provide a holistic understanding of the integration of vision-based intelligent systems in various aspects of precision agriculture. By providing a detailed discussion on key areas of digital life cycle of crops, this survey contributes to a deeper understanding of the complexities associated with the implementation of vision-guided intelligent systems in challenging agricultural environments. The focus of this survey is to explore widely used imaging and image analysis techniques being utilized for precision farming tasks. This paper first discusses various salient crop metrics used in digital agriculture. Then this paper illustrates the usage of imaging and computer vision techniques in various phases of digital life cycle of crops in precision agriculture, such as image acquisition, image stitching and photogrammetry, image analysis, decision making, treatment, and planning. After establishing a thorough understanding of related terms and techniques involved in the implementation of vision-based intelligent systems for precision agriculture, the survey concludes by outlining the challenges associated with implementing generalized computer vision models for real-time deployment of fully autonomous farms.

In this study, an Artificial Neural Network-Genetic Algorithm (ANN-GA) approach was successfully applied to optimise the physicochemical factors influencing the synthesis of unsaturated fatty acids (UFAs) in the microalgae P. kessleri UCM 001. The optimized model recommended specific cultivation conditions, including glucose at 29 g/L, NaNO₃ at 2.4 g/L, K₂HPO₄ at 0.4 g/L, red LED light, an intensity of 1000 lx, and an 8:16-h light-dark cycle. Through ANN-GA optimisation, a remarkable 66.79% increase in UFAs production in P. kessleri UCM 001 was achieved, compared to previous studies. This underscores the potential of this technology for enhancing valuable lipid production. Sequential variations in the application of physicochemical factors during microalgae culture under mixotrophic conditions, as optimized by ANN-GA, induced alterations in UFAs production and composition in P. kessleri UCM 001. This suggests the feasibility of tailoring the lipid profile of microalgae to obtain specific lipids for diverse industrial applications. The microalgae were isolated from a high-mountain lake in Colombia, highlighting their adaptation to extreme conditions. This underscores their potential for sustainable lipid and biomaterial production. This study demonstrates the effectiveness of using ANN-GA technology to optimise UFAs production in microalgae, offering a promising avenue for obtaining valuable lipids. The microalgae's unique origin in a high-mountain environment in Colombia emphasises the importance of exploring and harnessing microbial resources in distinctive geographical regions for biotechnological applications.

In recent years, smart agriculture has gained strength due to the application of industry 4.0 technologies in agriculture. As a result, efforts are increasing in proposing artificial vision applications to solve many problems. However, many of these applications are developed separately. Many academic works have proposed solutions integrating image classification techniques through IoT platforms. For this reason, this paper aims to answer the following research questions: (1) What are the main problems to be solved with smart farming IoT platforms that incorporate images? (2) What are the main strategies for incorporating image classification methods in smart agriculture IoT platforms? and (3) What are the main image acquisition, preprocessing, transmission, and classification technologies used in smart agriculture IoT platforms? This study adopts a Systematic Literature Review (SLR) approach. We searched Scopus, Web of Science, IEEE Xplore, and Springer Link databases from January 2018 to July 2022. From which we could identify five domains corresponding to (1) disease and pest detection, (2) crop growth and health monitoring, (3) irrigation and crop protection management, (4) intrusion detection, and (5) fruits and plant counting. There are three types of strategies to integrate image data into smart agriculture IoT platforms: (1) classification process in the edge, (2) classification process in the cloud, and (3) classification process combined. The main advantage of the first is obtaining data in real-time, and its main disadvantage is the cost of implementation. On the other hand, the main advantage of the second is the ability to process high-resolution images, and its main disadvantage is the need for high-bandwidth connectivity. Finally, the mixed strategy can significantly benefit infrastructure investment, but most works are experimental.

Estimation of damage in plants is a key issue for crop protection. Currently, experts in the field manually assess the plots. This is a time-consuming task that can be automated thanks to the latest technology in computer vision (CV). The use of image-based systems and recently deep learning-based systems have provided good results in several agricultural applications. These image-based applications outperform expert evaluation in controlled environments, and now they are being progressively included in non-controlled field applications.

A novel solution based on deep learning techniques in combination with image processing methods is proposed to tackle the estimate of plant damage in the field. The proposed solution is a two-stage algorithm. In a first stage, the single plants in the plots are detected by an object detection YOLO based model. Then a regression model is applied to estimate the damage of each individual plant. The solution has been developed and validated in oilseed rape plants to estimate the damage caused by flea beetle.

The crop detection model achieves a mean precision average of 91% with a [email protected] of 0.99 and a [email protected] of 0.91 for oilseed rape specifically. The regression model to estimate up to 60% of damage degree in single plants achieves a MAE of 7.11, and R2 of 0.46 in comparison with manual evaluations done plant by plant by experts. Models are deployed in a docker, and with a REST API communication protocol they can be inferred directly for images acquired in the field from a mobile device.

Plant disease detection has played a significant role in combating plant diseases that pose a threat to global agriculture and food security. Detecting these diseases early can help mitigate their impact and ensure healthy crop yields. Machine learning algorithms have emerged as powerful tools for accurately identifying and classifying a wide range of plant diseases from trained image datasets of affected crops. These algorithms, including deep learning algorithms, have shown remarkable success in recognizing disease patterns and early signs of plant diseases. Besides early detection, there are other potential benefits of machine learning algorithms in overall plant disease management, such as soil and climatic condition predictions for plants, pest identification, proximity detection, and many more. Over the years, research has focused on using machine-learning algorithms for plant disease detection. Nevertheless, little is known about the extent to which the research community has explored machine learning algorithms to cover other significant areas of plant disease management. In view of this, we present a cross-comparative review of machine learning algorithms and applications designed for plant disease detection with a specific focus on four (4) economically important plants: apple, cassava, cotton, and potato. We conducted a systematic review of articles published between 2013 and 2023 to explore trends in the research community over the years. After filtering a number of articles based on our inclusion criteria, including articles that present individual prediction accuracy for classes of disease associated with the selected plants, 113 articles were considered relevant. From these articles, we analyzed the state-of-the-art techniques, challenges, and future prospects of using machine learning for disease identification of the selected plants. Results from our review show that deep learning and other algorithms performed significantly well in detecting plant diseases. In addition, we found a few references to plant disease management covering prevention, diagnosis, control, and monitoring. In view of this, little or no work has explored the prediction of the recovery of affected plants. Hence, we propose opportunities for developing machine learning-based technologies to cover prevention, diagnosis, control, monitoring, and recovery.