Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9804165
Shaopeng Hu, Jiafeng Zhang, Zhiwu Li
This paper deals with the problem of diagnosability enforcement of discrete event systems. Given a non-diagnosable discrete event system modelled with Petri nets, which may enter a deadlock state or an unobservable live-lock (that each composed transition is unobservable), a digital twin system (derived from the original labeled Petri net model) can be established as a particular Petri net such that the original system under the control of its digital twin system is diagnosable. An example is given to illustrate the proposed method and the correctness of the method is proved by theoretical proof.
{"title":"Diagnosability Enforcement in Labeled Petri Nets Based on Digital Twins","authors":"Shaopeng Hu, Jiafeng Zhang, Zhiwu Li","doi":"10.1109/CoDIT55151.2022.9804165","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9804165","url":null,"abstract":"This paper deals with the problem of diagnosability enforcement of discrete event systems. Given a non-diagnosable discrete event system modelled with Petri nets, which may enter a deadlock state or an unobservable live-lock (that each composed transition is unobservable), a digital twin system (derived from the original labeled Petri net model) can be established as a particular Petri net such that the original system under the control of its digital twin system is diagnosable. An example is given to illustrate the proposed method and the correctness of the method is proved by theoretical proof.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116786023","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9803887
N. S. M. Sauki, N. S. Damanhuri, N. A. Othman, Y. Chiew, Belinda Chong Chiew Meng, M. Nor, Nurhidayah Mohd Zainol, A. Ralib
Most mathematical models were developed to guide clinicians in managing patients who are mechanically ventilated (MV) in intensive care unit (ICU). However, asynchrony events (AE) could occur when a patient's breathing is not synchronized with the MV support, which is caused by spontaneously breathing (SB) effort or mismatch of inspiratory and expiratory timings of ventilator support even though the patients are fully sedated. One of the real metrics that can detect AEs in MV patients is through time varying elastance estimation. Previous studies found that SB patients developed a negative elastance as a result of the SB effort put forth by these patients. Hence, this study aims to estimate the AEs of MV patients by adding negative elastance (AUC Edrs_negative) in the model. Data were obtained from nine mechanically ventilated respiratory failure patients from the International Islamic University Malaysia (IIUM) Hospital. Asynchrony index (AInew) represents a total estimation of AEs and the negative elastance in MV patients. Patients’ data were classified by ventilation mode, and AInew was computed for each of the patients and compared with the previous methods in calculating the AI. The results show that the new modelbased technique in estimating the value of AInew has produced a higher value as compared to previous measurements of AIori as expected. Hence, this new measurement of AI has successfully shown that by adding AEs and AUC Edrs negative together, this model is more sensitive and precisely measures the AI especially during the synchronized intermittent mandatory ventilation (SIMV) mode. Thus, the estimation of AEs with negative elastance may aid clinicians in selecting the appropriate MV ventilation mode and allow for precise respiratory mechanics monitoring, especially in SB patients.
{"title":"Estimation of Asynchrony Events with Negative Elastance in Spontaneously Breathing Mechanically Ventilated Patients in ICU","authors":"N. S. M. Sauki, N. S. Damanhuri, N. A. Othman, Y. Chiew, Belinda Chong Chiew Meng, M. Nor, Nurhidayah Mohd Zainol, A. Ralib","doi":"10.1109/CoDIT55151.2022.9803887","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9803887","url":null,"abstract":"Most mathematical models were developed to guide clinicians in managing patients who are mechanically ventilated (MV) in intensive care unit (ICU). However, asynchrony events (AE) could occur when a patient's breathing is not synchronized with the MV support, which is caused by spontaneously breathing (SB) effort or mismatch of inspiratory and expiratory timings of ventilator support even though the patients are fully sedated. One of the real metrics that can detect AEs in MV patients is through time varying elastance estimation. Previous studies found that SB patients developed a negative elastance as a result of the SB effort put forth by these patients. Hence, this study aims to estimate the AEs of MV patients by adding negative elastance (AUC Edrs_negative) in the model. Data were obtained from nine mechanically ventilated respiratory failure patients from the International Islamic University Malaysia (IIUM) Hospital. Asynchrony index (AInew) represents a total estimation of AEs and the negative elastance in MV patients. Patients’ data were classified by ventilation mode, and AInew was computed for each of the patients and compared with the previous methods in calculating the AI. The results show that the new modelbased technique in estimating the value of AInew has produced a higher value as compared to previous measurements of AIori as expected. Hence, this new measurement of AI has successfully shown that by adding AEs and AUC Edrs negative together, this model is more sensitive and precisely measures the AI especially during the synchronized intermittent mandatory ventilation (SIMV) mode. Thus, the estimation of AEs with negative elastance may aid clinicians in selecting the appropriate MV ventilation mode and allow for precise respiratory mechanics monitoring, especially in SB patients.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127212738","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9803908
I. Oluigbo, H. Seba, Mohammed Haddad
Learning low dimensional representations requires an expressive technique capable of capturing the different features for nodes, the relationship between nodes in the network and thus their similarities. However, many existing embedding techniques focus only on capturing the structural patterns in the network by randomly sampling the nodes in the neighborhood of the target node. To deal with this issue, we propose DSNCR, a node representation framework which uses the non-linear node attributes as well as their neighbourhood structural information to capture nodes similarities. This approach computes a semi-supervised regression analysis on the node attributes to guide a flexible probability walk procedure, such that different neighbourhoods are explored to capture rich network attributes and structures in a learned embedding. We verify the effectiveness of our model on link prediction and node classification tasks using real-life benchmark datasets, for which our technique performs better than existing embedding methods.
{"title":"Decision-based Sampling for Node Context Representation","authors":"I. Oluigbo, H. Seba, Mohammed Haddad","doi":"10.1109/CoDIT55151.2022.9803908","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9803908","url":null,"abstract":"Learning low dimensional representations requires an expressive technique capable of capturing the different features for nodes, the relationship between nodes in the network and thus their similarities. However, many existing embedding techniques focus only on capturing the structural patterns in the network by randomly sampling the nodes in the neighborhood of the target node. To deal with this issue, we propose DSNCR, a node representation framework which uses the non-linear node attributes as well as their neighbourhood structural information to capture nodes similarities. This approach computes a semi-supervised regression analysis on the node attributes to guide a flexible probability walk procedure, such that different neighbourhoods are explored to capture rich network attributes and structures in a learned embedding. We verify the effectiveness of our model on link prediction and node classification tasks using real-life benchmark datasets, for which our technique performs better than existing embedding methods.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"2410 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127478816","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9804035
R. Mallouhy, C. Guyeux, C. A. Jaoude, A. Makhoul
Fire brigades’ operations vary with time, climate, season, occasions, etc. For example, the frequency of accidents is greater during the day than at night. Thus, adjusting the need to the demand of fire departments by categories of operations can lead to a reduction of material, financial and human resources. Therefore, it can be very helpful during the financial and economic crisis most countries face. It also helps firefighters to be well prepared by knowing the type and number of human resources needed for the next operation. The aim of this study is to predict the number of firefighters’ interventions of 14 different categories varying between emergency and non-emergency deployments. The experiments in this study on the dataset provided by the fire and rescue service, SDIS 25, in the Doubs-France region showed that it is not necessary to improve the prediction when more explanatory variables are added. Some characteristics are not informative and may reduce the accuracy of the results.
{"title":"Machine Learning for Predicting Firefighters’ Interventions Per Type of Mission","authors":"R. Mallouhy, C. Guyeux, C. A. Jaoude, A. Makhoul","doi":"10.1109/CoDIT55151.2022.9804035","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9804035","url":null,"abstract":"Fire brigades’ operations vary with time, climate, season, occasions, etc. For example, the frequency of accidents is greater during the day than at night. Thus, adjusting the need to the demand of fire departments by categories of operations can lead to a reduction of material, financial and human resources. Therefore, it can be very helpful during the financial and economic crisis most countries face. It also helps firefighters to be well prepared by knowing the type and number of human resources needed for the next operation. The aim of this study is to predict the number of firefighters’ interventions of 14 different categories varying between emergency and non-emergency deployments. The experiments in this study on the dataset provided by the fire and rescue service, SDIS 25, in the Doubs-France region showed that it is not necessary to improve the prediction when more explanatory variables are added. Some characteristics are not informative and may reduce the accuracy of the results.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124901859","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9803942
G. Cherif, B. Trouillet, A. Toguyéni
This paper is about routing problem of Automated Intelligent Vehicles (AIVs) in a port. The objective is to prove the effectiveness of automated system for the transfer of containers in port terminals. The capacity of ships keeps increasing which means a huge volume of transported containers. For competition reasons, the time needed to load/unload ships must be reduced. For that, an automated system becomes a priority in order to increase the productivity and reduce operating costs. This paper is about modeling and scheduling problems to enable the routing of the AIV s. The modeling is done with transition timed Petri nets (T - TPN) that behave under earliest firing policy. The scheduling problem of AIVs is treated using Beam search with the obj ective of finding a control sequence between an initial state and a reference one with minimal time. An example is used to illustrate the approach.
{"title":"Modeling and routing problems of automated port using T-TPN and Beam search","authors":"G. Cherif, B. Trouillet, A. Toguyéni","doi":"10.1109/CoDIT55151.2022.9803942","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9803942","url":null,"abstract":"This paper is about routing problem of Automated Intelligent Vehicles (AIVs) in a port. The objective is to prove the effectiveness of automated system for the transfer of containers in port terminals. The capacity of ships keeps increasing which means a huge volume of transported containers. For competition reasons, the time needed to load/unload ships must be reduced. For that, an automated system becomes a priority in order to increase the productivity and reduce operating costs. This paper is about modeling and scheduling problems to enable the routing of the AIV s. The modeling is done with transition timed Petri nets (T - TPN) that behave under earliest firing policy. The scheduling problem of AIVs is treated using Beam search with the obj ective of finding a control sequence between an initial state and a reference one with minimal time. An example is used to illustrate the approach.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"628 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123276619","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9804122
Mohamed Mouaici
Road traffic safety is a major concern for road authorities and ordinary citizens. Consequently, accident prediction has become an interesting research topic that tries to provide solutions to predict, in real-time, traffic accidents occurrence and their severity. In this paper, a meta-model to predict, as early as possible, the risk and the severity of traffic accidents is proposed. The meta-model exploits seven predictive algorithms widely used in the literature and relies on a majority voting mechanism to improve predictions. For this purpose, a dataset of more than 45,000 observations is used, and two accident levels are considered. Based on the localization and the time of each accident, non-accident data are generated to create negative observations in the final dataset. Moreover, several features related to traffic flow, weather, and road conditions are collected and used as predictors to build and evaluate the predictive solutions and the meta-model. The experiment results show that the proposed meta-model dominates all other models in terms of F1 score.
{"title":"Traffic Accident Severity Prediction Using a Meta-Model Based on a Majority Vote","authors":"Mohamed Mouaici","doi":"10.1109/CoDIT55151.2022.9804122","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9804122","url":null,"abstract":"Road traffic safety is a major concern for road authorities and ordinary citizens. Consequently, accident prediction has become an interesting research topic that tries to provide solutions to predict, in real-time, traffic accidents occurrence and their severity. In this paper, a meta-model to predict, as early as possible, the risk and the severity of traffic accidents is proposed. The meta-model exploits seven predictive algorithms widely used in the literature and relies on a majority voting mechanism to improve predictions. For this purpose, a dataset of more than 45,000 observations is used, and two accident levels are considered. Based on the localization and the time of each accident, non-accident data are generated to create negative observations in the final dataset. Moreover, several features related to traffic flow, weather, and road conditions are collected and used as predictors to build and evaluate the predictive solutions and the meta-model. The experiment results show that the proposed meta-model dominates all other models in terms of F1 score.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"14 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123788063","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9803973
E. M. Barhoumi, P. Okonkwo, Manaf Zghaibeh, I. B. Belgacem, Slah Farhani, F. Bacha
The development of green hydrogen production processes is highly praised due to its advantages in environment protection and economy growth. Within this framework, this paper discusses the feasibility of hydrogen production in grid connected photovoltaic power station. The system is based on the conversion of solar energy into electrical power and hydrogen gas. The produced electrical energy is converted into hydrogen through electrolysis process. Therefore, the produced hydrogen will be used for refueling fuel cell vehicles or any other applications. The paper discusses the economic efficiency and the optimization of the PV grid connected system for the production of hydrogen. The total capacity of the grid connected PV system is 1500 kWp. The system produces 36465 kg of hydrogen per year with an average of 400 kg/day. The Levelized cost of hydrogen production is 4.2 €/kg. The estimated cost of the project depends on the cost of civil works, PV systems, electrolysers, and hydrogen tanks.
{"title":"Optimization of PV-Grid Connected System Based Hydrogen Refueling Station","authors":"E. M. Barhoumi, P. Okonkwo, Manaf Zghaibeh, I. B. Belgacem, Slah Farhani, F. Bacha","doi":"10.1109/CoDIT55151.2022.9803973","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9803973","url":null,"abstract":"The development of green hydrogen production processes is highly praised due to its advantages in environment protection and economy growth. Within this framework, this paper discusses the feasibility of hydrogen production in grid connected photovoltaic power station. The system is based on the conversion of solar energy into electrical power and hydrogen gas. The produced electrical energy is converted into hydrogen through electrolysis process. Therefore, the produced hydrogen will be used for refueling fuel cell vehicles or any other applications. The paper discusses the economic efficiency and the optimization of the PV grid connected system for the production of hydrogen. The total capacity of the grid connected PV system is 1500 kWp. The system produces 36465 kg of hydrogen per year with an average of 400 kg/day. The Levelized cost of hydrogen production is 4.2 €/kg. The estimated cost of the project depends on the cost of civil works, PV systems, electrolysers, and hydrogen tanks.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125434050","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9804043
F. Moutai, Mohammed Amine Tajioue, Salma Azzouzi, M. E. H. Charaf
The emergence of the Internet of Things (IoT) has raised several issues related to the development and deployment of IoT applications within IT infrastructures. In this context, cloud computing is often faced with latency issues when hosting such applications, even though the cloud environment offers promising opportunities to increase productivity and reduce costs significantly. Therefore, it is important to review the current conformance scheme in the cloud environments by considering the coordination challenges during data processing. For this purpose, we suggest an architecture for conformance testing of IoT based implementations in the Cloud. The idea is to set parallel testers to handle the conformance of the distributed implementation with respect to the specification. In this case, the testing process must support coordination between the different distributed components in order to detect the resulting faults. Therefore, the main contribution of this work is to propose a new architecture based on Markov decision processes with an adaptive controller to monitor and optimize the overall testing process in the distributed Cloud.
{"title":"An Optimal Approach for Testing Control in The Distributed Cloud","authors":"F. Moutai, Mohammed Amine Tajioue, Salma Azzouzi, M. E. H. Charaf","doi":"10.1109/CoDIT55151.2022.9804043","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9804043","url":null,"abstract":"The emergence of the Internet of Things (IoT) has raised several issues related to the development and deployment of IoT applications within IT infrastructures. In this context, cloud computing is often faced with latency issues when hosting such applications, even though the cloud environment offers promising opportunities to increase productivity and reduce costs significantly. Therefore, it is important to review the current conformance scheme in the cloud environments by considering the coordination challenges during data processing. For this purpose, we suggest an architecture for conformance testing of IoT based implementations in the Cloud. The idea is to set parallel testers to handle the conformance of the distributed implementation with respect to the specification. In this case, the testing process must support coordination between the different distributed components in order to detect the resulting faults. Therefore, the main contribution of this work is to propose a new architecture based on Markov decision processes with an adaptive controller to monitor and optimize the overall testing process in the distributed Cloud.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125506912","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9804003
R. Fellag, F. Yacef, M. Guiatni, M. Hamerlain, Laid Degaa, N. Rizoug
The study of control techniques for exoskeleton robots used in upper-extremity rehabilitation is gaining popularity. These robots are connected to the human upper limb at multiple points, allowing for smooth and independent joint movements. Therefore, providing a robust, precise, and safe control system is necessary. Sliding control-based approaches are well reputed for their robustness to parameter uncertainties, modeling errors, and external disturbances. Nevertheless, their major disadvantage is chattering. This paper describes two controllers based on sliding mode theory to reduce this undesirable effect: the generalized variable structure control and the higher-order finite-time sliding mode control. The former incorporates the derivative of the torque input in the model, while the latter is based on homogeneity and higher-order sliding modes to diminish chattering. A comparison between the two controllers is achieved by simulating passive rehabilitation mode.
{"title":"Upper Limb Exoskeleton Robot Control Using Input Output Switching","authors":"R. Fellag, F. Yacef, M. Guiatni, M. Hamerlain, Laid Degaa, N. Rizoug","doi":"10.1109/CoDIT55151.2022.9804003","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9804003","url":null,"abstract":"The study of control techniques for exoskeleton robots used in upper-extremity rehabilitation is gaining popularity. These robots are connected to the human upper limb at multiple points, allowing for smooth and independent joint movements. Therefore, providing a robust, precise, and safe control system is necessary. Sliding control-based approaches are well reputed for their robustness to parameter uncertainties, modeling errors, and external disturbances. Nevertheless, their major disadvantage is chattering. This paper describes two controllers based on sliding mode theory to reduce this undesirable effect: the generalized variable structure control and the higher-order finite-time sliding mode control. The former incorporates the derivative of the torque input in the model, while the latter is based on homogeneity and higher-order sliding modes to diminish chattering. A comparison between the two controllers is achieved by simulating passive rehabilitation mode.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125563538","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}
Pub Date : 2022-05-17DOI: 10.1109/CoDIT55151.2022.9803899
A. Sava, L. Ichim, D. Popescu
The paper's goal was to create some neural networks-based models for the detection and classification of insects such as Halyomorpha Halys in ecological orchards, from acquired images in the trees. The detecting operations were performed using models from two of the most efficient deep learning families in this area: R-CNN and YOLO. Using the proposed models, (Faster R-CNN, YOLOv5-s, YOLOv5-m, and YOLOv5-1) to early detection of harmful insects, a real contribution to anticipating damage in orchards is possible. The dataset is composed of images taken from the Maryland Biodiversity dataset. All training and testing operations were performed with the help of GPU processors provided by Google, the resulting models being saved on Google Drive Cloud. The images were evaluated from the detection and the classification perspective based on specific metrics such as precision, recall, and mAP. The best results were obtained for YOLOv5-m.
{"title":"Detection of Halyomorpha Halys Using Neural Networks","authors":"A. Sava, L. Ichim, D. Popescu","doi":"10.1109/CoDIT55151.2022.9803899","DOIUrl":"https://doi.org/10.1109/CoDIT55151.2022.9803899","url":null,"abstract":"The paper's goal was to create some neural networks-based models for the detection and classification of insects such as Halyomorpha Halys in ecological orchards, from acquired images in the trees. The detecting operations were performed using models from two of the most efficient deep learning families in this area: R-CNN and YOLO. Using the proposed models, (Faster R-CNN, YOLOv5-s, YOLOv5-m, and YOLOv5-1) to early detection of harmful insects, a real contribution to anticipating damage in orchards is possible. The dataset is composed of images taken from the Maryland Biodiversity dataset. All training and testing operations were performed with the help of GPU processors provided by Google, the resulting models being saved on Google Drive Cloud. The images were evaluated from the detection and the classification perspective based on specific metrics such as precision, recall, and mAP. The best results were obtained for YOLOv5-m.","PeriodicalId":185510,"journal":{"name":"2022 8th International Conference on Control, Decision and Information Technologies (CoDIT)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124709894","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：