Computers and Electronics in Agriculture最新文献

{"title":"RGB camera-based monocular stereo vision applied in plant phenotype: A survey","authors":"","doi":"10.1016/j.compag.2024.109523","DOIUrl":"10.1016/j.compag.2024.109523","url":null,"abstract":"<div><h3>Background</h3><div>The breeding of plants with superior traits and the improvement of cultivation means are two essential ways to achieve yield growth and quality promotion. Phenotype, which is the result of the interaction between genes and the environment, plays a key role in understanding plant geometry, growth and development. However, inefficient manual phenotypic measurement has become the main bottleneck restricting the advancement of related technologies. The monocular stereo vision system based on an RGB camera is considered as a promising approach for achieving high-throughput three-dimensional phenotypic data acquisition. This approach is cost-effective, highly efficient, and accurate.</div></div><div><h3>Scope and approach</h3><div>This work presents a comprehensive summary of the eight commonly used three-dimensional reconstruction methods in monocular stereo vision, along with three common image acquisition methods (circular, fixed, and straight) applied in plant phenotyping. Through a systematic review of the literature published in the past decade, this paper highlights the application of these systems and matching methods in three-dimensional plant phenotypic research. Additionally, this paper provides a discussion on the advantages and disadvantages of different approaches.</div></div><div><h3>Key findings and conclusions</h3><div>At present, monocular stereo vision systems based on a single RGB camera are widely utilized to acquire diverse plant traits due to their affordability and convenience. Different application scenarios have corresponding mechanical structure and data processing methods. Deep learning-based three-dimensional reconstruction methods have demonstrated promising results and significant potential across all three common image acquisition methods. However, the current effectiveness of deep learning in reconstruction requires further validation in the absence of datasets. Moreover, limitations exist in utilizing the results of 3D reconstruction and in the selection of experimental subjects, such as vertical farming. To advance modern breeding and intelligent cultivation, it is imperative to promote dataset collection, diversify the range of research subjects (such as edible fungi and diseased plants), and develop a novel, automated, high-throughput, four-dimensional phenotype platform. As such, monocular stereo vision systems based on an RGB camera, coupled with expanded applications and the development of more efficient reconstruction algorithms, will undoubtedly emerge as a focal point for future researches.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"LVF: A language and vision fusion framework for tomato diseases segmentation","authors":"","doi":"10.1016/j.compag.2024.109484","DOIUrl":"10.1016/j.compag.2024.109484","url":null,"abstract":"<div><div>With the development of deep learning technology, the control of tomato diseases has emerged as a crucial aspect of intelligent agricultural management. While current research on tomato disease segmentation has made considerable strides, challenges persist due to the susceptibility of tomato leaf diseases to strong light reflections and shadow gradients in sunlight. Additionally, the complex backgrounds found in agricultural fields often lead to model confusion, resulting in inaccurate segmentation. Traditional methods for tomato disease segmentation rely on single-modal image-based models, which struggle when dealing with the nuanced features and limited scope of tomato leaf diseases. To address these issues, our study introduces the LVF framework, a dual-modal approach combining image and text information for pre-segmentation of tomato diseases. We began by creating a new dataset labeled with both images and text, specifically focusing on diseased tomato leaves with guidance from agricultural experts. For image processing, we developed a probabilistic differential fusion network to mitigate interference caused by high-frequency noise, leveraging color and grayscale images. Furthermore, our reinforcement feature network and threshold filtering network enhance useful information while filtering out negative information from the fused images. In text processing, we proposed a multi-scale cross-nesting network to integrate semantic information about diseases across different scales and types. By nesting Bert-processed word vectors with fused image vectors, our model gains a deeper understanding of semantic information, thereby improving its ability to segment crop diseases accurately. Our experiments, conducted on self-constructed tomato datasets as well as public datasets for tomatoes and maize, demonstrated the efficacy and robustness of our approach in leaf disease segmentation. The LVF framework offers a valuable tool to enhance the accuracy of crop disease segmentation, especially in complex agricultural environments.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-feature language-image model for fruit quality image classification","authors":"","doi":"10.1016/j.compag.2024.109462","DOIUrl":"10.1016/j.compag.2024.109462","url":null,"abstract":"<div><div>Fruit quality classification has a great impact on the modern fruit industry. However, deep learning methods for fruit quality classification often demand a substantial number of labeled samples, which are hard and expensive to collect in many real-world applications, resulting in overfitting and low generalization. The Contrastive Language-Image Pre-Training (CLIP) model, which fuses image and text features, has demonstrated excellent performance in zero-shot classification. Inspired by CLIP, in this paper, we propose a multi-feature language-image (MFLI) model for fruit quality classification, where the fruit image and feature text are fused to enhance feature extraction. Furthermore, we construct a pomelo quality dataset containing first- and second-grade pomelo. Based on the zero-shot learning results of CLIP on this dataset, we provide recommendations for pre-prompt and multi-feature text. Experimental results show that in both zero-shot, few-shot, and conventional learning sceneries, our MFLI model outperforms state-of-the-art models on seven types of fruits, demonstrating excellent generalization capabilities.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418429","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cost-efficient algorithm for autonomous cultivators: Implementing template matching with field digital twins for precision agriculture","authors":"","doi":"10.1016/j.compag.2024.109509","DOIUrl":"10.1016/j.compag.2024.109509","url":null,"abstract":"<div><div>The paper focuses on the development of a vision system to automate the position control of a cultivator used for crop weeding. The vision algorithm allows monitoring of the cultivator’s misalignment with respect to crop rows, with real-time processing. The key content includes the introduction of a self-generated digital twin of the field model for numerical validation of different computer vision solutions and a comparison of three vision algorithms for measuring deviation. The objectives of the study are to improve the precision of misalignment measurements and ensure safe and accurate movement of the cultivator. The rationale behind the study is to address constraints such as camera installation and crop color, and to emphasize the importance of a confidence estimation feature for accurate measurement. The paper also provides an overview of related works in the literature, highlighting the two phases of plant identification and deviation measurement. Tests carried out on soybean and maize crops demonstrate the improvements allowed by the proposed algorithm in terms of higher measurement precision, even in the presence of high weed infestation or a significant number of missing plants. Additionally, the paper suggests analysis simplifications to enhance the algorithm’s speed while maintaining satisfactory measurement accuracy.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418339","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accelerated Data Engine: A faster dataset construction workflow for computer vision applications in commercial livestock farms","authors":"","doi":"10.1016/j.compag.2024.109452","DOIUrl":"10.1016/j.compag.2024.109452","url":null,"abstract":"<div><div>Large-scale, high-quality dataset was the foundation of developing advanced artificial intelligence applications. However, creating such a benchmark dataset in a professional field, such as precision management of animals, was always a challenge because of the costly and labor-intensive process of annotation and review. This study introduced a novel workflow named Accelerated Data Engine (ADE), designed to efficiently produce representative and high-quality computer vision datasets from raw animal surveillance footage. By incorporating referring and grounding models (R&amp;G models) as auto-annotators, along with a distillation mechanism for dataset-auditors, ADE significantly speeded up the dataset construction process. The new workflow received natural language inputs as referrals to identify animal instances, delineated their body shapes, and then refined the auto-annotated data through a selection process. To demonstrate the efficacy of ADE, three 30-minute surveillance video samples featuring pigs, sheep, and cattle were discussed in this study. The results indicated the R&amp;G models effectively annotated animals across various farms, while distillation mechanisms could identify various detection errors, balance the data representations, refine annotations, and verify the data quality. Two high-quality cattle datasets (6.5 k and 486 frames), including 26 k and 2.5 k cattle instances, were generated through the ADE workflow from 24-hour surveillance videos on a commercial cattle farm and made publicly available. The proposed dataset has achievable performance between 74.6 %∼84.1 %. The ADE workflow saved 78.4 % of manual work compared to the traditional dataset construction workflow (approximately 141 h). This pioneering approach empowered the fast creation of benchmark animal datasets and would enhance computer vision applications in the livestock production industry in the future.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142423276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Modeling and control of dissolved oxygen in recirculating aquaculture systems: A circadian rhythm analysis approach and GSMPC controller","authors":"","doi":"10.1016/j.compag.2024.109515","DOIUrl":"10.1016/j.compag.2024.109515","url":null,"abstract":"<div><div>Precise control of dissolved oxygen (DO) concentration is significant for the growth and development of aquatic products. This study focused on modeling and control DO concentration in a recirculating aquaculture system (RAS). Firstly, a DO dynamic model was established based on the oxygen mass transfer equation and circadian rhythm of fish oxygen consumption rate. The fitting R² of simulation data and measured data of DO response were both above 0.96 and the significance of circadian rhythm in DO dynamic model was confirmed. Subsequently, Gain-scheduling model predictive control (GSMPC) suitable for circadian rhythm was proposed and applied to regulate the DO concentration in fish tank under various operating points, and its performance was compared with that of traditional model predictive control (MPC). In terms of set value tracking, Integral of Absolute Error of GSMPC controller dropped by 23.46% compared to MPC controller, and Integral Squared Error dropped by 11.27%, while for energy consumption, the Integral of Absolute Control dropped by 11.28%. These results demonstrated GSMPC controller not only reduced the error but also shrank the energy consumption. The findings highlighted the notable advantages of GSMPC over traditional MPC, emphasizing its effectiveness in precisely regulating DO concentration in a RAS based on circadian rhythm.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142417970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of an online prediction system for soil organic matter and soil moisture content based on multi-modal fusion","authors":"","doi":"10.1016/j.compag.2024.109514","DOIUrl":"10.1016/j.compag.2024.109514","url":null,"abstract":"<div><div>Realizing accurate determination of in situ soil organic matter (SOM) content and soil moisture (SM) content in the field is of great importance for improving agricultural production efficiency. However, the feature information of a single sensor is limited, and there is still a certain gap in modeling accuracy compared to traditional laboratory methods. The variations of SM in the field also interfere with data collection, limiting the application of single sensor. In order to realize the efficient detection of in-situ SOM and SM content, this study developed an online detection system for SOM and SM content based on the fusion of characteristic wavelengths, visible images and thermal imaging image features. Firstly, based on a vehicle-mounted platform, a visible-thermal imaging camera and a characteristic wavelength integration device were integrated in a deep pine plow to realize the simultaneous acquisition of in-situ soil multi-sensor data. Then, a lightweight multimodal network was constructed to obtain thermal visible light image features and characteristic wavelength features through branch networks, achieving deep fusion of different modal data and predicting SOM and SM content. Finally, output the forecasting results of SOM and SM content, and transmit the predictive information to the cloud platform for storage. It was verified that the Multi-modal proposed in this study worked best in the laboratory environment, with a predicted R² of 0.91 and RMSE of 2.9 g/kg for SOM, and a predicted R² of 0.92 and RMSE of 0.77 % for SM. Compared with single image or characteristic spectral data, the fusion of visible image and characteristic wavelength effectively improved the prediction accuracy of SOM. The real-time prediction of SM was realized by fusing thermal imaging data, and the elimination of the moisture effect of visible images and characteristic wavelengths was also realized with the help of deep fusion of Multi-modal network. After field validation, the R² of the Multi-modal system was 0.84 and the RMSE was 5.0 g/kg for SOM, and the R² of the SM was 0.88 and the RMSE was 1.03 %. Despite the differences in soil types between the validation field and the sampling field, the system still demonstrated strong generalization and achieved high accuracy prediction of in-situ SOM and SM in the field, which effectively improves the efficiency and applicability of field. The efficiency and applicability of soil testing effectively improve the efficiency and provide technical guidance for precision management in the field.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418345","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generation of NIR Spectral Band from RGB Image with Wavelet Domain Spectral Extrapolation Generative Adversarial Network","authors":"","doi":"10.1016/j.compag.2024.109461","DOIUrl":"10.1016/j.compag.2024.109461","url":null,"abstract":"<div><div>Near-infrared (NIR) imaging exhibits outstanding penetration capabilities and robust anti-interference characteristics. However, acquiring high-resolution and high-fidelity NIR images is difficult due to the limited mobility, high cost, and low resolution of NIR imaging hardware. In this paper, we proposed a new end-to-end Wavelet Domain Spectral Extrapolation Generative Adversarial Network (WSEGAN) to generate highly realistic NIR images from RGB images. Since RGB and NIR images have different types of noise and artifacts, which affects the quality of NIR images generation. We design a generator with a discrete wavelet transform and an attention mechanism to capture multi-resolution contextual information, and a multi-scale discriminator to capture detailed and global features. Then, the proposed approach is evaluated by three different datasets to achieve optimal results regarding both visual effects and quantitative evaluation. The normalized difference vegetation index (NDVI) is utilized to validate the effectiveness of the generated NIR images. The result demonstrates a strong correlation between the generated images and the actual vegetation distribution. More importantly, the proposed network is testified to generate NIR images to be fused with the RGB image source for agricultural target detection tasks. In dark conditions, results show that using multi-modal data instead of RGB images improves <span><math><mrow><mi>m</mi><mi>A</mi><msub><mrow><mi>P</mi></mrow><mrow><mn>0</mn><mo>.</mo><mn>5</mn></mrow></msub></mrow></math></span> detection accuracy by 4% on the CAPSICUM dataset and by 8% on the KIWI dataset across five object detection methods. This follows the physical significance of the NIR imaging and demonstrates the potential use of the NIR images generated by the proposed method.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142417775","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Portable multiplexed ion-selective sensor for long-term and continuous irrigation water quality monitoring","authors":"","doi":"10.1016/j.compag.2024.109455","DOIUrl":"10.1016/j.compag.2024.109455","url":null,"abstract":"<div><div>In agricultural contexts, the demand for continuous and precise measurement of multiple ions is crucial. While arrays of solid-contact ion-selective electrodes (SCEs) have been developed previously, there has been limited emphasis on their continuous and long-term monitoring of ions. Addressing this gap, our work introduces an innovative sensor array utilizing Ni-HAB MOF as an ion-to-electron transducer, enabling real-time detection of nitrate, potassium, and pH levels. The sensors exhibit exceptional stability, eliminating the need for frequent recalibration. For instance, the K⁺-selective sensor displays an unprecedentedly low potential drift of 0.05 µV/h, surpassing existing solid-contact sensors by two orders of magnitude. Similarly, the pH sensor demonstrates a drift of 0.3 µV/h, outperforming competitors by a factor of 100. The NO₃^--selective sensor shows minimal drift at 0.5 µV/h, surpassing comparable sensors by a factor of ten. Additionally, the K⁺-selective sensor features a sensitivity of 57.8 mV/dec and a LOD of 1.9 µM, while the NO₃^--selective sensor offers a sensitivity of 56.8 mV/dec and a LOD of 6.23 µM. Integrated into a portable array with wireless data transmission, this system enables real-time water quality monitoring in remote areas. Rigorous testing of the developed sensor array in a tailored complex agricultural solution confirms its selective response to target ions even in the presence of interfering ions. Importantly, pH fluctuations do not compromise the precision of the K⁺ and NO₃^-- selective sensors, highlighting the system’s robustness in real-world agricultural settings.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418342","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Technologies in cattle traceability: A bibliometric analysis","authors":"","doi":"10.1016/j.compag.2024.109459","DOIUrl":"10.1016/j.compag.2024.109459","url":null,"abstract":"<div><div>It has been widely documented that livestock cattle can play a non-negligible role in natural landscapes due to climate change, deforestation, and enteric methane emissions. Alternatively, sustainable protocols and market digitalization are highlighted as promising tools to mitigate environmental cattle impacts by authenticated data in digital traceability systems. Digital inclusion, particularly for cattle breeders, can be a useful starting point for employing sustainable protocol in food chain production and management. This study analyzes the evolution of knowledge in the area of animal traceability to compare applied technologies found by a bibliometric analysis of articles published in Web of Science. The study evidences a clear change in thematic research over decades, currently culminating in technologies such as blockchain, IoT (Internet of Things), machine learning, and deep learning. These technologies emerge as the main research scopes in promoting transparency and reliability in the production chain, especially considering individual digital identification. However, challenges such as high investment requirements and difficulties in data accessibility, interoperability, privacy, and security implicate the low maturity level of available technologies and knowledge, therefore preventing further adoption and development of reliable worldwide animal traceability systems.</div></div>","PeriodicalId":50627,"journal":{"name":"Computers and Electronics in Agriculture","volume":null,"pages":null},"PeriodicalIF":7.7,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}