Information about dairy cow ruminating is closely associated with the health status of dairy cows. Therefore, it is of great significance to recognize and make statistics of dairy cows’ ruminating and feeding behavior. Concerning conventional recognition methods which are dependent on contact type devices, they have some defects of poor instantaneity and strong stress responses. As for recognition based on machine vision, it needs to transmit masses of data and raises high requirements for the cloud server and network performance. According to principles of edge computing, the model is deployed via Tensorflow.js in an edge device in the present study, constructing a recognition and statistical system for ruminating and feeding behavior of dairy cows. Through the application programming interface (API) of the browser, an edge device is able to invoke a camera and acquire dairy cow images. Then, the images can be inputted in the SSD MobileNet V2 model, which is followed by inference based on browser hashrate. Moreover, the edge device merely uploads recognition results to the cloud server for statistics, which features high instantaneity and compatibility. In terms of recognizing ruminating and feeding behavior of dairy cows, the proposed system has a precision ratio of 96.50%, a recall rate of 91.77%, an F1-score of 94.08%, specificity of 91.36%, and accuracy of 91.66%. This suggests that the proposed method is effective in recognizing dairy cow behavior.
{"title":"Recognition and statistical method of cows rumination and eating behaviors based on Tensorflow.js","authors":"Yu Zhang, Xiangting Li, Zhiqing Yang, Shaopeng Hu, Xiao Fu, Weizheng Shen","doi":"10.1016/j.inpa.2023.11.002","DOIUrl":"10.1016/j.inpa.2023.11.002","url":null,"abstract":"<div><div>Information about dairy cow ruminating is closely associated with the health status of dairy cows. Therefore, it is of great significance to recognize and make statistics of dairy cows’ ruminating and feeding behavior. Concerning conventional recognition methods which are dependent on contact type devices, they have some defects of poor instantaneity and strong stress responses. As for recognition based on machine vision, it needs to transmit masses of data and raises high requirements for the cloud server and network performance. According to principles of edge computing, the model is deployed via Tensorflow.js in an edge device in the present study, constructing a recognition and statistical system for ruminating and feeding behavior of dairy cows. Through the application programming interface (API) of the browser, an edge device is able to invoke a camera and acquire dairy cow images. Then, the images can be inputted in the SSD MobileNet V2 model, which is followed by inference based on browser hashrate. Moreover, the edge device merely uploads recognition results to the cloud server for statistics, which features high instantaneity and compatibility. In terms of recognizing ruminating and feeding behavior of dairy cows, the proposed system has a precision ratio of 96.50%, a recall rate of 91.77%, an F1-score of 94.08%, specificity of 91.36%, and accuracy of 91.66%. This suggests that the proposed method is effective in recognizing dairy cow behavior.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 581-589"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135514893","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.11.003
Jiawen Pan, Caicong Wu, Weixin Zhai
Field-road trajectory segmentation (FRTS) is a critical step in the processing of agricultural machinery trajectory data. This study presents a generalized optimization framework based on metaheuristic algorithms (MAs) to increase the accuracy of the field-road trajectory segmentation model. The MA optimization process is used in this framework to precisely and quickly identify the parameters of the FRTS model. It is difficult to solve the parameter optimization problem with basic metaheuristic algorithms without falling into local optima due to their insufficient performance. This study therefore combines a genetic algorithm (GA) with a slime mould algorithm (SMA) to propose a novel enhanced hybrid algorithm (GASMA); the algorithm has superior global search capability due to the implicit parallelism of the GA, and the oscillation concentration mechanism of the SMA is used to enhance the algorithm's local search capability. To maintain the balance between the two capacities, a nonlinear parameter management technique is developed that adaptively modifies the algorithm's computational process based on the fitness distribution deviation of the population. Experiments were conducted on real agricultural trajectory datasets with various sample frequencies, and the proposed algorithm was compared with existing methods to validate its efficiency. According to the experimental data, the optimized model produced better results. The proposed approach provides an automatic and accurate method for determining the optimal parameter configurations of FRTS model instances, where the parameter optimization solution is not confined to a single specified procedure and can be addressed by a variety of metaheuristic algorithms.
{"title":"A hybrid genetic slime mould algorithm for parameter optimization of field-road trajectory segmentation models","authors":"Jiawen Pan, Caicong Wu, Weixin Zhai","doi":"10.1016/j.inpa.2023.11.003","DOIUrl":"10.1016/j.inpa.2023.11.003","url":null,"abstract":"<div><div>Field-road trajectory segmentation (FRTS) is a critical step in the processing of agricultural machinery trajectory data. This study presents a generalized optimization framework based on metaheuristic algorithms (MAs) to increase the accuracy of the field-road trajectory segmentation model. The MA optimization process is used in this framework to precisely and quickly identify the parameters of the FRTS model. It is difficult to solve the parameter optimization problem with basic metaheuristic algorithms without falling into local optima due to their insufficient performance. This study therefore combines a genetic algorithm (GA) with a slime mould algorithm (SMA) to propose a novel enhanced hybrid algorithm (GASMA); the algorithm has superior global search capability due to the implicit parallelism of the GA, and the oscillation concentration mechanism of the SMA is used to enhance the algorithm's local search capability. To maintain the balance between the two capacities, a nonlinear parameter management technique is developed that adaptively modifies the algorithm's computational process based on the fitness distribution deviation of the population. Experiments were conducted on real agricultural trajectory datasets with various sample frequencies, and the proposed algorithm was compared with existing methods to validate its efficiency. According to the experimental data, the optimized model produced better results. The proposed approach provides an automatic and accurate method for determining the optimal parameter configurations of FRTS model instances, where the parameter optimization solution is not confined to a single specified procedure and can be addressed by a variety of metaheuristic algorithms.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 590-602"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139299152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.08.006
Fernando Henrique Lermen , Vera Lúcia Milani Martins , Marcia Elisa Echeveste , Filipe Ribeiro , Carla Beatriz da Luz Peralta , José Luis Duarte Ribeiro
This study assesses the value capture of a result-oriented Product-Service System offer that constitutes a post-harvest solution. Applying the reinforcement learning reward system and general linear models, we identified the Brazilian farmer's propensities to choose different products and services from the proposed system. Reinforcement learning enables one to understand the choice process by rewarding the attributes selected and applying penalties to those not chosen. Regarding product options, farmers' most valued attributes were extended capacity, fixed installation, automatic dryer, and CO2 emission control, considering the investigated system. Regarding service options, the farmers opted for maintenance plans, performance reports, no photovoltaic energy, and purchase over the rental modality. These results assist managers through a reward learning system that constantly updates the value assigned by farmers to product and service attributes. They allow real-time visualization of changes in farmers' preferences regarding the product-service system configurations.
{"title":"Reinforcement Learning system to capture value from Brazilian post-harvest offers","authors":"Fernando Henrique Lermen , Vera Lúcia Milani Martins , Marcia Elisa Echeveste , Filipe Ribeiro , Carla Beatriz da Luz Peralta , José Luis Duarte Ribeiro","doi":"10.1016/j.inpa.2023.08.006","DOIUrl":"10.1016/j.inpa.2023.08.006","url":null,"abstract":"<div><div>This study assesses the value capture of a result-oriented Product-Service System offer that constitutes a post-harvest solution. Applying the reinforcement learning reward system and general linear models, we identified the Brazilian farmer's propensities to choose different products and services from the proposed system. Reinforcement learning enables one to understand the choice process by rewarding the attributes selected and applying penalties to those not chosen. Regarding product options, farmers' most valued attributes were extended capacity, fixed installation, automatic dryer, and CO<sub>2</sub> emission control, considering the investigated system. Regarding service options, the farmers opted for maintenance plans, performance reports, no photovoltaic energy, and purchase over the rental modality. These results assist managers through a reward learning system that constantly updates the value assigned by farmers to product and service attributes. They allow real-time visualization of changes in farmers' preferences regarding the product-service system configurations.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 499-511"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42681215","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.06.001
Yongjun Zhang , Xinqing Xiao
Live fish waterless transportation could be recognized as an essential supplement for water-based transportation due to its low oxygen consumption and less waste water pollution. The critical problem to maintaining the fish survival quality under such a unique transport strategy is accurately controlling the oxygen concentration in the container to be constantly at stable and high levels. This paper aims to propose an improved fuzzy PID control system based on the grey model with residual rectification by improved particle swarm optimized Gated Recurrent Unit (GM-IPSO-GRU) to realize advanced oxygen level control. In addition, it is also reinforced by adopting the improved grey wolf optimization (IGWO) for the majorization of control parameters (quantization factors, scale factors) with full consideration of fish size features. In this study, Turbot (Scophthalmus maximus) is taken as the test subject to verify the integrated control performance of the optimized fuzzy PID controller through simulated waterless live transportation under low-temperature conditions. The proposed control system is validated as more efficient than the traditional proportional integral derivative (PID) and fuzzy PID algorithms for handling its nonlinear, time-varying, and time lag problems well. In summary, the control group experiment shows that the newly-designed control system has the advantages of shorter stabilization time, minor overshoot, and strong anti-interference ability for oxygen level adjustment. Finally, applying this novel control technology can effectively improve oxygen adjustment efficiency and provide feasible quality control support for the deep optimization of the live fish circulation industry.
{"title":"Fuzzy PID control system optimization and verification for oxygen-supplying management in live fish waterless transportation","authors":"Yongjun Zhang , Xinqing Xiao","doi":"10.1016/j.inpa.2023.06.001","DOIUrl":"10.1016/j.inpa.2023.06.001","url":null,"abstract":"<div><div>Live fish waterless transportation could be recognized as an essential supplement for water-based transportation due to its low oxygen consumption and less waste water pollution. The critical problem to maintaining the fish survival quality under such a unique transport strategy is accurately controlling the oxygen concentration in the container to be constantly at stable and high levels. This paper aims to propose an improved fuzzy PID control system based on the grey model with residual rectification by improved particle swarm optimized Gated Recurrent Unit (GM-IPSO-GRU) to realize advanced oxygen level control. In addition, it is also reinforced by adopting the improved grey wolf optimization (IGWO) for the majorization of control parameters (quantization factors, scale factors) with full consideration of fish size features. In this study, Turbot (Scophthalmus maximus) is taken as the test subject to verify the integrated control performance of the optimized fuzzy PID controller through simulated waterless live transportation under low-temperature conditions. The proposed control system is validated as more efficient than the traditional proportional integral derivative (PID) and fuzzy PID algorithms for handling its nonlinear, time-varying, and time lag problems well. In summary, the control group experiment shows that the newly-designed control system has the advantages of shorter stabilization time, minor overshoot, and strong anti-interference ability for oxygen level adjustment. Finally, applying this novel control technology can effectively improve oxygen adjustment efficiency and provide feasible quality control support for the deep optimization of the live fish circulation industry.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 421-437"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42570678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.09.001
Lingling Yang, Xingshi Xu, Jizheng Zhao, Huaibo Song
Cows’ posture change is the fatal influencing factor for accurate identification of individual cows. To achieve non-contact, high-precision detection and identification of individual cows in farm environment, a cow individual identification method by the fusion of RetinaFace and improved FaceNet was proposed. MobileNet-enhanced RetinaFace was applied to ameliorate the impact of output channel quantity and convolution kernel dynamics using depthwise convolution combined with pointwise convolution. Regression predictions of bovine facial features and keypoints were generated under varying distances, scales and sizes. FaceNet's core feature network was enhanced through MobileNet integration, and the loss function was jointly optimized with Cross Entropy Loss and Triplet Loss to achieve a quicker and more stable convergence curve. The distances between the generated embedding vectors of cow facial features were corresponding to the similarity between cow faces, enabling accurate matching. RetinaFace exhibited detection false negative rates of 2.67%, 0.66%, 2.67%, and 3.33% under conditions of occlusion, no occlusion, low light, and bright light for cow facial detection. For cow facial pattern detection, the false negative rates for black and white patterns, pure black and pure white were 1.33%, 6.00% and 8.00%, respectively. Regarding cow facial posture changes, the false negative rates for face upward, bowing down, profile, and normal posture were 1.33%, 1.33%, 4.00% and 0.66%, respectively. Improved FaceNet model achieved an accuray of 99.50% on training set and 83.60% on test set. In comparison to YOLOX, the recognition model presented in this research demonstrated increased accuracy in cow facial detection under occlusion, no occlusion and strong lighting conditions by 2.67%, 0.40%, and 0.40%, respectively. Moreover, the accuracy for patterns with pure black and pure white tones surpassed that of YOLOX by 1.06% and 5.71%, correspondingly. Additionally, the accuracy rates for face upward, bowing down, profile and normal posture were higher than YOLOX by 2.00%, 3.34%, 2.66% and 0.40%, respectively. The proposed model demonstrates the proficiency in accurately identifying individual cows in natural scenes.
{"title":"Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes","authors":"Lingling Yang, Xingshi Xu, Jizheng Zhao, Huaibo Song","doi":"10.1016/j.inpa.2023.09.001","DOIUrl":"10.1016/j.inpa.2023.09.001","url":null,"abstract":"<div><div>Cows’ posture change is the fatal influencing factor for accurate identification of individual cows. To achieve non-contact, high-precision detection and identification of individual cows in farm environment, a cow individual identification method by the fusion of RetinaFace and improved FaceNet was proposed. MobileNet-enhanced RetinaFace was applied to ameliorate the impact of output channel quantity and convolution kernel dynamics using depthwise convolution combined with pointwise convolution. Regression predictions of bovine facial features and keypoints were generated under varying distances, scales and sizes. FaceNet's core feature network was enhanced through MobileNet integration, and the loss function was jointly optimized with Cross Entropy Loss and Triplet Loss to achieve a quicker and more stable convergence curve. The distances between the generated embedding vectors of cow facial features were corresponding to the similarity between cow faces, enabling accurate matching. RetinaFace exhibited detection false negative rates of 2.67%, 0.66%, 2.67%, and 3.33% under conditions of occlusion, no occlusion, low light, and bright light for cow facial detection. For cow facial pattern detection, the false negative rates for black and white patterns, pure black and pure white were 1.33%, 6.00% and 8.00%, respectively. Regarding cow facial posture changes, the false negative rates for face upward, bowing down, profile, and normal posture were 1.33%, 1.33%, 4.00% and 0.66%, respectively. Improved FaceNet model achieved an accuray of 99.50% on training set and 83.60% on test set. In comparison to YOLOX, the recognition model presented in this research demonstrated increased accuracy in cow facial detection under occlusion, no occlusion and strong lighting conditions by 2.67%, 0.40%, and 0.40%, respectively. Moreover, the accuracy for patterns with pure black and pure white tones surpassed that of YOLOX by 1.06% and 5.71%, correspondingly. Additionally, the accuracy rates for face upward, bowing down, profile and normal posture were higher than YOLOX by 2.00%, 3.34%, 2.66% and 0.40%, respectively. The proposed model demonstrates the proficiency in accurately identifying individual cows in natural scenes.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 512-523"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42302982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.08.001
Dan Xu , Yanfeng Li , Anguo Dai , Shumei Zhao , Weitang Song
Greenhouse climate is crucial for crop growth. Traditional climate control techniques are carried out through on–off actuators based on growers’ experience. Advanced control algorithms usually track setpoints through continuous control inputs. These setpoints cannot guarantee maximum profit, which can be treated as the control objective of the optimal control algorithm. This paper investigated on–off optimal control algorithms based on two-time-scale decomposition. Mixed-integer nonlinear dynamic programming is used in the fast subproblem to quantify the influence of restricting different control inputs to be integers on the control objective and the CPU time. Results show that compared with continuous control inputs, a decrease of 2.21 ¥·m−2 in the control objective and an increase of 7.84·103 s in the CPU time can be found when defining all control inputs to be integers with 12 collocation points in one day. The methods of sorting and pulse width modulation are used to simulate the receding horizon optimal control in the whole growing period. Results show that compared with continuous control inputs, decreases of 83.54 ¥·m−2 and 4.45 ¥·m−2 can be found with the methods of sorting and pulse width modulation. Moreover, the method of pulse width modulation cannot guarantee state constraint satisfaction. This paper suggests modifying actuators to supply continuous control inputs before implementing optimal control algorithms for maximum profit.
{"title":"Model-based quantitative analysis in two-time-scale decomposed on–off optimal control of greenhouse cultivation","authors":"Dan Xu , Yanfeng Li , Anguo Dai , Shumei Zhao , Weitang Song","doi":"10.1016/j.inpa.2023.08.001","DOIUrl":"10.1016/j.inpa.2023.08.001","url":null,"abstract":"<div><div>Greenhouse climate is crucial for crop growth. Traditional climate control techniques are carried out through on–off actuators based on growers’ experience. Advanced control algorithms usually track setpoints through continuous control inputs. These setpoints cannot guarantee maximum profit, which can be treated as the control objective of the optimal control algorithm. This paper investigated on–off optimal control algorithms based on two-time-scale decomposition. Mixed-integer nonlinear dynamic programming is used in the fast subproblem to quantify the influence of restricting different control inputs to be integers on the control objective and the CPU time. Results show that compared with continuous control inputs, a decrease of 2.21 ¥·m<sup>−2</sup> in the control objective and an increase of 7.84·10<sup>3</sup> s in the CPU time can be found when defining all control inputs to be integers with 12 collocation points in one day. The methods of sorting and pulse width modulation are used to simulate the receding horizon optimal control in the whole growing period. Results show that compared with continuous control inputs, decreases of 83.54 ¥·m<sup>−2</sup> and 4.45 ¥·m<sup>−2</sup> can be found with the methods of sorting and pulse width modulation. Moreover, the method of pulse width modulation cannot guarantee state constraint satisfaction. This paper suggests modifying actuators to supply continuous control inputs before implementing optimal control algorithms for maximum profit.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 488-498"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48146231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.09.002
Santoshi Rudrakar, Parag Rughani
IoT based agriculture (Ag-IoT) is an emerging communication technology that is widely adopted by agricultural entrepreneurs and farmers to perform agricultural agro-chores in the farm to improve productivity, for better monitoring, and to reduce labor costs. However, the use of the Internet in Ag-IoT facilitates real-time functionality in an agriculture system, it can increase the risk of security breaches and cyber attacks that would cause the Ag-IoT system to malfunction and can affect its productivity. Ag-IoT is overlooked in cyber security parameters, which can have severe impacts on its trustworthiness and adoption by agricultural communities. To address this gap, this article presents a systematic study of the literature published between 2001 and 2023 that discusses advances in Ag-IoT technology. The subjects included in the study on Ag-IoT are emerging applications, different IoT architectures, suspected cyber attacks and cyber crimes, and challenges in incident response and digital forensics. The findings of this study encourage the reader to explore future potential research avenues related to the security risks and challenges of Ag-IoT, as well as the readiness for incident response and forensic investigation in the smart agricultural sector. The main conclusion of this study is that security must be ensured in Ag-IoT environments to offer uninterrupted services and also there is a need for forensic readiness for effective investigation in the event of unanticipated security incidents.
{"title":"IoT based Agriculture (Ag-IoT): A detailed study on Architecture, Security and Forensics","authors":"Santoshi Rudrakar, Parag Rughani","doi":"10.1016/j.inpa.2023.09.002","DOIUrl":"10.1016/j.inpa.2023.09.002","url":null,"abstract":"<div><div>IoT based agriculture (Ag-IoT) is an emerging communication technology that is widely adopted by agricultural entrepreneurs and farmers to perform agricultural agro-chores in the farm to improve productivity, for better monitoring, and to reduce labor costs. However, the use of the Internet in Ag-IoT facilitates real-time functionality in an agriculture system, it can increase the risk of security breaches and cyber attacks that would cause the Ag-IoT system to malfunction and can affect its productivity. Ag-IoT is overlooked in cyber security parameters, which can have severe impacts on its trustworthiness and adoption by agricultural communities. To address this gap, this article presents a systematic study of the literature published between 2001 and 2023 that discusses advances in Ag-IoT technology. The subjects included in the study on Ag-IoT are emerging applications, different IoT architectures, suspected cyber attacks and cyber crimes, and challenges in incident response and digital forensics. The findings of this study encourage the reader to explore future potential research avenues related to the security risks and challenges of Ag-IoT, as well as the readiness for incident response and forensic investigation in the smart agricultural sector. The main conclusion of this study is that security must be ensured in Ag-IoT environments to offer uninterrupted services and also there is a need for forensic readiness for effective investigation in the event of unanticipated security incidents.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 524-541"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42340537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01DOI: 10.1016/j.inpa.2023.10.001
Keyang Zhong , Xueqian Sun , Gedi Liu , Yifeng Jiang , Yi Ouyang , Yang Wang
Environmental parameter data collected by sensors for monitoring the environment of agricultural facility operations are usually incomplete due to external environmental disturbances and device failures. And the missing of collected data is completely at random. In practice, missing data could create biased estimations and make multivariate time series predictions of environmental parameters difficult, leading to imprecise environmental control. A multivariate time series imputation model based on generative adversarial networks and multi-head attention (ATTN-GAN) is proposed in this work to reducing the negative consequence of missing data. ATTN-GAN can capture the temporal and spatial correlation of time series, and has a good capacity to learn data distribution. In the downstream experiments, we used ATTN-GAN and baseline models for data imputation, and predicted the imputed data, respectively. For the imputation of missing data, over the 20%, 50% and 80% missing rate, ATTN-GAN had the lowest RMSE, 0.1593, 0.2012 and 0.2688 respectively. For water temperature prediction, data processed with ATTN-GAN over MLP, LSTM, DA-RNN prediction methods had the lowest MSE, 0.6816, 0.8375 and 0.3736 respectively. Those results revealed that ATTN-GAN outperformed all baseline models in terms of data imputation accuracy. The data processed by ATTN-GAN is the best for time series prediction.
{"title":"Attention-based generative adversarial networks for aquaponics environment time series data imputation","authors":"Keyang Zhong , Xueqian Sun , Gedi Liu , Yifeng Jiang , Yi Ouyang , Yang Wang","doi":"10.1016/j.inpa.2023.10.001","DOIUrl":"10.1016/j.inpa.2023.10.001","url":null,"abstract":"<div><div>Environmental parameter data collected by sensors for monitoring the environment of agricultural facility operations are usually incomplete due to external environmental disturbances and device failures. And the missing of collected data is completely at random. In practice, missing data could create biased estimations and make multivariate time series predictions of environmental parameters difficult, leading to imprecise environmental control. A multivariate time series imputation model based on generative adversarial networks and multi-head attention (ATTN-GAN) is proposed in this work to reducing the negative consequence of missing data. ATTN-GAN can capture the temporal and spatial correlation of time series, and has a good capacity to learn data distribution. In the downstream experiments, we used ATTN-GAN and baseline models for data imputation, and predicted the imputed data, respectively. For the imputation of missing data, over the 20%, 50% and 80% missing rate, ATTN-GAN had the lowest RMSE, 0.1593, 0.2012 and 0.2688 respectively. For water temperature prediction, data processed with ATTN-GAN over MLP, LSTM, DA-RNN prediction methods had the lowest MSE, 0.6816, 0.8375 and 0.3736 respectively. Those results revealed that ATTN-GAN outperformed all baseline models in terms of data imputation accuracy. The data processed by ATTN-GAN is the best for time series prediction.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 542-551"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136009786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work investigated the potential of metal oxide semiconductor (MOS) gas sensors for environmental monitoring of methane. Calibrations were performed under controlled conditions in the lab, and under semi-controlled conditions in the field, using a modified head space chamber set-up. Concentrations up to ±300 ppm methane were tested. The relationship between sensor conductance and methane concentrations could be very well described using principles from adsorption theory. The adjustable parameters were background conductance G0, a sensitivity constant S and a non-ideality coefficient n, where n has a non-rational value between 0 and 1. Sensor behaviour was very different in dry air than in humid air, with the background conductance increasing approximately tenfold and sensitivity decreasing between 20 fold and 80 fold, while the non-ideality coefficient increased from ±0.4 to ±0.6. Nevertheless, at high methane concentrations comparable conductance values were recorded in dry and humid air. The standard deviation of predicted values was 1.6 μS.for the least well described dataset. Using the corresponding calibration curve, a detection limit of 11 ppm is calculated for humid ambient air. This values suggests that MOS sensor are adequately sensitive to be used for methane detection in an agricultural context.
{"title":"Evaluation of the applicability of a metal oxide semiconductor gas sensor for methane emissions from agriculture","authors":"Bastiaan Molleman , Enrico Alessi , Fabio Passaniti , Karen Daly","doi":"10.1016/j.inpa.2023.11.001","DOIUrl":"10.1016/j.inpa.2023.11.001","url":null,"abstract":"<div><div>This work investigated the potential of metal oxide semiconductor (MOS) gas sensors for environmental monitoring of methane. Calibrations were performed under controlled conditions in the lab, and under semi-controlled conditions in the field, using a modified head space chamber set-up. Concentrations up to ±300 ppm methane were tested. The relationship between sensor conductance and methane concentrations could be very well described using principles from adsorption theory. The adjustable parameters were background conductance G<sub>0</sub>, a sensitivity constant S and a non-ideality coefficient n, where n has a non-rational value between 0 and 1. Sensor behaviour was very different in dry air than in humid air, with the background conductance increasing approximately tenfold and sensitivity decreasing between 20 fold and 80 fold, while the non-ideality coefficient increased from ±0.4 to ±0.6. Nevertheless, at high methane concentrations comparable conductance values were recorded in dry and humid air. The standard deviation of predicted values was 1.6 μS.for the least well described dataset. Using the corresponding calibration curve, a detection limit of 11 ppm is calculated for humid ambient air. This values suggests that MOS sensor are adequately sensitive to be used for methane detection in an agricultural context.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"11 4","pages":"Pages 573-580"},"PeriodicalIF":7.7,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135454758","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-23DOI: 10.1016/j.inpa.2024.09.004
Uzair Sajjad , Farzana Bibi , Imtiyaz Hussain , Naseem Abbas , Muhammad Sultan , Hafiz Muhammad Asfahan , Muhammad Aleem , Wei-Mon Yan
Drying technologies have been essential for extending the shelf-life of perishable fruits and vegetables for over a century. Vacuum freeze-drying (VFD), though invented over a hundred years ago, remains one of the most advanced drying techniques, known for sustainably drying perishable products while maintaining quality indices and morphological properties comparable to their fresh state. The performance of the VFD system is sensitive to the operating conditions and features of the drying product which is assessed using experimental and/or numerical methods. However, the qualitative aspects of the dried product are not predictable. In this context, the present study aims to create a deep neural framework (DNF) that predicts the performance of a Vacuum Freeze Drying (VFD) system for kiwifruit, based on its morphology and nutritional value under varying conditions. This involves translating the fruit’s morphological features into trainable data and using a Generative Adversarial Network (GAN) to create diverse, unlabeled datasets. The framework is optimized using Gaussian Process (GP) for hyper-parameter tuning, focusing on minimizing errors like mean square error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). The maximum MSE of 1.243 is found in the prediction of rehydration rate, followed by color (0.725), energy consumption (0.426), moisture content (0.379), texture (0.320), sensory (0.250), and Brix (0.215), respectively. The maximum MAE and MAPE values are recorded 0.833 and 32.99 % while the minimum is observed 0.368 and 7.019 % in the case of rehydration rate and Brix, respectively. Overall, the R2 value was computed 0.863 which is reasonable for the quality assessment of kiwifruit dried by the VFD system.
{"title":"Machine learning enabled assessment of the vacuum freeze-drying of the kiwifruit","authors":"Uzair Sajjad , Farzana Bibi , Imtiyaz Hussain , Naseem Abbas , Muhammad Sultan , Hafiz Muhammad Asfahan , Muhammad Aleem , Wei-Mon Yan","doi":"10.1016/j.inpa.2024.09.004","DOIUrl":"10.1016/j.inpa.2024.09.004","url":null,"abstract":"<div><div>Drying technologies have been essential for extending the shelf-life of perishable fruits and vegetables for over a century. Vacuum freeze-drying (VFD), though invented over a hundred years ago, remains one of the most advanced drying techniques, known for sustainably drying perishable products while maintaining quality indices and morphological properties comparable to their fresh state. The performance of the VFD system is sensitive to the operating conditions and features of the drying product which is assessed using experimental and/or numerical methods. However, the qualitative aspects of the dried product are not predictable. In this context, the present study aims to create a deep neural framework (DNF) that predicts the performance of a Vacuum Freeze Drying (VFD) system for kiwifruit, based on its morphology and nutritional value under varying conditions. This involves translating the fruit’s morphological features into trainable data and using a Generative Adversarial Network (GAN) to create diverse, unlabeled datasets. The framework is optimized using Gaussian Process (GP) for hyper-parameter tuning, focusing on minimizing errors like mean square error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). The maximum MSE of 1.243 is found in the prediction of rehydration rate, followed by color (0.725), energy consumption (0.426), moisture content (0.379), texture (0.320), sensory (0.250), and Brix (0.215), respectively. The maximum MAE and MAPE values are recorded 0.833 and 32.99 % while the minimum is observed 0.368 and 7.019 % in the case of rehydration rate and Brix, respectively. Overall, the R<sup>2</sup> value was computed 0.863 which is reasonable for the quality assessment of kiwifruit dried by the VFD system.</div></div>","PeriodicalId":53443,"journal":{"name":"Information Processing in Agriculture","volume":"12 2","pages":"Pages 245-259"},"PeriodicalIF":7.7,"publicationDate":"2024-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144115864","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号