Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220520
A. Periola, K. Ogudo, A. Alonge
The stratosphere is an aeronautical resource whose use is of benefit to the government in delivering aviation services. It also provides a freely cooling environment making it suitable for hosting non-terrestrial data centers. However, the development of a framework enabling the utilization of the stratosphere requires further research attention. The research presents a multientity architecture that describes the role of a stratosphere-bound airport that supports the deployment and use of future stratosphere-based data centers. The solution being presented is intended to increase the operational duration of future deployed stratosphere-based data centers. The focus here is on enhancing the operational duration of the stratosphere-based data center. This is important for its role in future networks. Analysis shows that the proposed solution improved the operational duration by at least 33% and by up to 76% on average.
{"title":"Realizing the Potential of Stratosphere Utilization via Stratosphere Data Centers","authors":"A. Periola, K. Ogudo, A. Alonge","doi":"10.1109/icABCD59051.2023.10220520","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220520","url":null,"abstract":"The stratosphere is an aeronautical resource whose use is of benefit to the government in delivering aviation services. It also provides a freely cooling environment making it suitable for hosting non-terrestrial data centers. However, the development of a framework enabling the utilization of the stratosphere requires further research attention. The research presents a multientity architecture that describes the role of a stratosphere-bound airport that supports the deployment and use of future stratosphere-based data centers. The solution being presented is intended to increase the operational duration of future deployed stratosphere-based data centers. The focus here is on enhancing the operational duration of the stratosphere-based data center. This is important for its role in future networks. Analysis shows that the proposed solution improved the operational duration by at least 33% and by up to 76% on average.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"4 1","pages":"1-6"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80701322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220494
M. Okwu, I. Emovon, O. J. Oyejide, Kingsley C. Ezekiel, Olaye Messiah, Perpetua C. Jones-Iwuagwu
Automated Guided Vehicles (AGVs) are widely used as delivery agents and for material transportation in factories, hospital environment, and other facilities. Conducting performance tests on AGVs has the potential to ratify and improve the efficiency, and reliability of the system. However, published studies on performance analysis focused on classical metrics for such evaluation. In this study, the emphasis is on the performance evaluation of a developed lightweight AGV using the Adaptive Neuro-fuzzy Inference System (ANFIS). The developed line following AGV is flexible, intelligent, and nifty, and can be accessed wirelessly, and controlled by an operator. It was programmed to avoid collision with the help of a proximity sensor attached. The performance test was conducted by drawing black lines on a plain surface for easy navigation of the AGV. A series of experiments was carried out by using realistic test variables like the navigation pattern of AGV, test accuracy, energy efficiency, obstacle avoidance, task accomplishment, and others. Sensitivity analysis was done using the ANFIS surface plot. The total system intelligence (TSI) obtained for the different trials are 76%; 79%; 80%; 81%; 79% and 81 %, for the first, second, third, fourth, fifth, and final trials respectively. The preeminent observable performance was the fourth and sixth trials, obtained at 81 %. The outcome of the investigation reveals that the ANFIS model is an efficient soft computing technique capable of performing TSI tests of AGVs with a high degree of accuracy. The model is also recommended in AGV platooning.
{"title":"Performance Analysis of a Light Weight Ground Robotic Vehicle by Implementing Adaptive Neuro-Fuzzy Inference System (ANFIS)","authors":"M. Okwu, I. Emovon, O. J. Oyejide, Kingsley C. Ezekiel, Olaye Messiah, Perpetua C. Jones-Iwuagwu","doi":"10.1109/icABCD59051.2023.10220494","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220494","url":null,"abstract":"Automated Guided Vehicles (AGVs) are widely used as delivery agents and for material transportation in factories, hospital environment, and other facilities. Conducting performance tests on AGVs has the potential to ratify and improve the efficiency, and reliability of the system. However, published studies on performance analysis focused on classical metrics for such evaluation. In this study, the emphasis is on the performance evaluation of a developed lightweight AGV using the Adaptive Neuro-fuzzy Inference System (ANFIS). The developed line following AGV is flexible, intelligent, and nifty, and can be accessed wirelessly, and controlled by an operator. It was programmed to avoid collision with the help of a proximity sensor attached. The performance test was conducted by drawing black lines on a plain surface for easy navigation of the AGV. A series of experiments was carried out by using realistic test variables like the navigation pattern of AGV, test accuracy, energy efficiency, obstacle avoidance, task accomplishment, and others. Sensitivity analysis was done using the ANFIS surface plot. The total system intelligence (TSI) obtained for the different trials are 76%; 79%; 80%; 81%; 79% and 81 %, for the first, second, third, fourth, fifth, and final trials respectively. The preeminent observable performance was the fourth and sixth trials, obtained at 81 %. The outcome of the investigation reveals that the ANFIS model is an efficient soft computing technique capable of performing TSI tests of AGVs with a high degree of accuracy. The model is also recommended in AGV platooning.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"52 1","pages":"1-7"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85157357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220515
Bhanu Prakash Reddy Banda, Bianca Govan, K. Roy, Kelvin S. Bryant
Malware attacks have become a crucial problem in modern life. From 2015 to 2021 about 56.1billion malware attacks have taken place in the world. A malware attack typically costs a business over 2.5 million dollars to remediate. According to Cybersecurity Ventures, during the next five years, the cost of cybercrime would increase by 15% yearly, reaching 10.5 trillion USD annually by 2025 from 3 trillion USD in 2015. There is a global epidemic of malware. Studies imply that malware's effects are deteriorating. The main defense against malware tools is malware detectors. Therefore, it is crucial that we research malware detection methods to better comprehend their advantages and disadvantages. This research focuses on an Application Pro-gramming Interface (API) call-based malware detection strategy with Machine Learning to further improve malware detection. The Limitations that motivated to work on this project was the lack of datasets with newly attacked malware samples and also lack of detecting the malware with good accuracy. The main goal of this research is to understand the malware behavior on the Windows platform, use a dynamic analysis to identify various aspects or features that have dangerous code patterns from malware samples and employ various malware and benign samples to construct and validate machine learning-based malware detection models. The data was gathered from publicly accessible sites and sampled using a sandbox approach. Machine Learning models were built using the new dataset. The Supervised Learning models and deep Learning models were applied to the dataset and then the results were compared and cross-checked to get the best fit model. This investigation demonstrated the possibility of estab- lishing a high-precision capability for the detection of malware while combining API calls and Machine Learning models., The strategy yielded a high malware detection accuracy of 88.26% (XGBoost) model and 90.70% (MLP classifier) for Windows-based platforms. We have used Explainable Machine Learning, namely the SHapley Additive exPlanations (SHAP) value methods to demonstrate the important component or feature responsible for the prediction of the model.
{"title":"Malware detection using Explainable ML models based on Feature Extraction using API calls","authors":"Bhanu Prakash Reddy Banda, Bianca Govan, K. Roy, Kelvin S. Bryant","doi":"10.1109/icABCD59051.2023.10220515","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220515","url":null,"abstract":"Malware attacks have become a crucial problem in modern life. From 2015 to 2021 about 56.1billion malware attacks have taken place in the world. A malware attack typically costs a business over 2.5 million dollars to remediate. According to Cybersecurity Ventures, during the next five years, the cost of cybercrime would increase by 15% yearly, reaching 10.5 trillion USD annually by 2025 from 3 trillion USD in 2015. There is a global epidemic of malware. Studies imply that malware's effects are deteriorating. The main defense against malware tools is malware detectors. Therefore, it is crucial that we research malware detection methods to better comprehend their advantages and disadvantages. This research focuses on an Application Pro-gramming Interface (API) call-based malware detection strategy with Machine Learning to further improve malware detection. The Limitations that motivated to work on this project was the lack of datasets with newly attacked malware samples and also lack of detecting the malware with good accuracy. The main goal of this research is to understand the malware behavior on the Windows platform, use a dynamic analysis to identify various aspects or features that have dangerous code patterns from malware samples and employ various malware and benign samples to construct and validate machine learning-based malware detection models. The data was gathered from publicly accessible sites and sampled using a sandbox approach. Machine Learning models were built using the new dataset. The Supervised Learning models and deep Learning models were applied to the dataset and then the results were compared and cross-checked to get the best fit model. This investigation demonstrated the possibility of estab- lishing a high-precision capability for the detection of malware while combining API calls and Machine Learning models., The strategy yielded a high malware detection accuracy of 88.26% (XGBoost) model and 90.70% (MLP classifier) for Windows-based platforms. We have used Explainable Machine Learning, namely the SHapley Additive exPlanations (SHAP) value methods to demonstrate the important component or feature responsible for the prediction of the model.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"3 1","pages":"1-7"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85206168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220490
Funmilayo S. Moninuola, E. Adetiba, Anthony A. Atayero, A. Awelewa, A. Adeyeye, Oluwadamilola Oshin, J. Ameh, A. Abayomi, Victor Ezekiel
Lung Cancer (LC), have the highest mortality rate and the second-highest incidence rate of all cancers combined because of a pathophysiological imbalance in the fundamental mechanism of cell proliferation. For patients with LC, prompt diagnosis and treatment are of utmost importance. The orthodox methods employed for detecting LC are characterised by invasiveness, protracted duration, high cost and exhibit reduced efficacy in detecting malignant cells during the initial phases of the ailment. The increasing attention of researchers toward the potential of utilising Volatile Organic Compound (VOC) biomarkers for the non-invasive detection of LC can be attributed to the advancements in techniques and procedures. This study offers a state-of-the-art portable E-nose that has the potential to enhance clinical outcomes associated with the early diagnosis of LC. Three ML models - SVM, AdaBoost, and MLP were employed to discriminate LC from other respiratory breathprint dataset. The MLP model achieved the highest performance accuracy result of 89.05%, specificity 95.12%, and sensitivity of 80%.
{"title":"Early Detection of Lung Cancer via Breath Analysis Utilising Electronic Nose","authors":"Funmilayo S. Moninuola, E. Adetiba, Anthony A. Atayero, A. Awelewa, A. Adeyeye, Oluwadamilola Oshin, J. Ameh, A. Abayomi, Victor Ezekiel","doi":"10.1109/icABCD59051.2023.10220490","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220490","url":null,"abstract":"Lung Cancer (LC), have the highest mortality rate and the second-highest incidence rate of all cancers combined because of a pathophysiological imbalance in the fundamental mechanism of cell proliferation. For patients with LC, prompt diagnosis and treatment are of utmost importance. The orthodox methods employed for detecting LC are characterised by invasiveness, protracted duration, high cost and exhibit reduced efficacy in detecting malignant cells during the initial phases of the ailment. The increasing attention of researchers toward the potential of utilising Volatile Organic Compound (VOC) biomarkers for the non-invasive detection of LC can be attributed to the advancements in techniques and procedures. This study offers a state-of-the-art portable E-nose that has the potential to enhance clinical outcomes associated with the early diagnosis of LC. Three ML models - SVM, AdaBoost, and MLP were employed to discriminate LC from other respiratory breathprint dataset. The MLP model achieved the highest performance accuracy result of 89.05%, specificity 95.12%, and sensitivity of 80%.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"9 1","pages":"1-6"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82427584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220464
Lukoki Mpova, T. Shongwe, Ali N. Hasan
The classification and detection of cyanosis using in-vivo and in-silico image processing approaches are intriguing and very special. In this study, a peripheral and central cyanosis image classification approach, using lightweight-deep learning Convolutional Neural Networks (CNNs), referred to as pre-trained MobileNet architecture, was introduced. This modified MobileNet model was assessed using the sanctioned dataset of 1300-image collected from multiple cyanosis published datasets. The augmentation technique was applied on the training dataset to enrich the productivity. Emphatic results, validation-accuracy and accuracies on the training and test datasets of 95% and 97%, respectively; were obtained as compared to the validation-accuracy of 79% and 82% of the Simple Convolutional Neural Networks (SCNNs) and Fine-tuned VGG16 models attained from prior stud.
{"title":"Classification and Detection of Cyanosis Images on Lightly and Darkly Pigmented Individual Human Skins using a Fine-Tuned MobileNet Architecture","authors":"Lukoki Mpova, T. Shongwe, Ali N. Hasan","doi":"10.1109/icABCD59051.2023.10220464","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220464","url":null,"abstract":"The classification and detection of cyanosis using in-vivo and in-silico image processing approaches are intriguing and very special. In this study, a peripheral and central cyanosis image classification approach, using lightweight-deep learning Convolutional Neural Networks (CNNs), referred to as pre-trained MobileNet architecture, was introduced. This modified MobileNet model was assessed using the sanctioned dataset of 1300-image collected from multiple cyanosis published datasets. The augmentation technique was applied on the training dataset to enrich the productivity. Emphatic results, validation-accuracy and accuracies on the training and test datasets of 95% and 97%, respectively; were obtained as compared to the validation-accuracy of 79% and 82% of the Simple Convolutional Neural Networks (SCNNs) and Fine-tuned VGG16 models attained from prior stud.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"41 1","pages":"1-5"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82493027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220508
Alden Boby, Dane Brown, James Connan, Marc Marais, Luxulo Lethukuthula Kuhlane
Licence plate recognition has many practical applications for security and surveillance. This paper presents a robust licence plate detection system that uses string-matching algorithms to identify a vehicle in data. Object detection models have had limited application in the character recognition domain. The system utilises the YOLO object detection model to perform character recognition to ensure more accurate character predictions. The model incorporates super-resolution techniques to enhance the quality of licence plate images to increase character recognition accuracy. The proposed system can accurately detect license plates in diverse conditions and can handle license plates with varying fonts and backgrounds. The system's effectiveness is demonstrated through experimentation on components of the system, showing promising license plate detection and character recognition accuracy. The overall system works with all the components to track vehicles by matching a target string with detected licence plates in a scene. The system has potential applications in law enforcement, traffic management, and parking systems and can significantly advance surveillance and security through automation.
{"title":"Enabling Vehicle Search Through Robust Licence Plate Detection","authors":"Alden Boby, Dane Brown, James Connan, Marc Marais, Luxulo Lethukuthula Kuhlane","doi":"10.1109/icABCD59051.2023.10220508","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220508","url":null,"abstract":"Licence plate recognition has many practical applications for security and surveillance. This paper presents a robust licence plate detection system that uses string-matching algorithms to identify a vehicle in data. Object detection models have had limited application in the character recognition domain. The system utilises the YOLO object detection model to perform character recognition to ensure more accurate character predictions. The model incorporates super-resolution techniques to enhance the quality of licence plate images to increase character recognition accuracy. The proposed system can accurately detect license plates in diverse conditions and can handle license plates with varying fonts and backgrounds. The system's effectiveness is demonstrated through experimentation on components of the system, showing promising license plate detection and character recognition accuracy. The overall system works with all the components to track vehicles by matching a target string with detected licence plates in a scene. The system has potential applications in law enforcement, traffic management, and parking systems and can significantly advance surveillance and security through automation.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"99 1","pages":"1-7"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78251777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220569
Ramahlapane Lerato Moila, M. Velempini
This study proposes an optimised routing scheme, called OCS-AODV, for Cognitive Radio Ad Hoc Networks (CRAHNs) to enhance Quality of Service (QoS). The scheme applies the Cuckoo Search (CS) algorithm optimised with a fitness function to improve the performance of the Ad Hoc On-Demand Distance Vector (AODV). The objective of the study is to evaluate the proposed scheme's performance with respect to delay, packet loss, packet delivery ratio and throughput. The literature review shows that the existing routing protocols have limitations which impact performance in dynamic environments. The proposed OCS-AODV scheme aims to address these limitations by selecting reliable paths based on a fitness function that considers the lifetime of nodes, reliability, and available buffer capacity. The simulation results have shown that the OCS-AODV scheme outperforms the CS-DSDV and ACO-AODV schemes in terms of PDR, packet loss, delay, and throughput. The study concludes that the proposed scheme improves the QoS of routing in CRAHNs. However, the use of a single fitness function may not be optimal for all network scenarios. Multiple fitness functions may be considered in future and the schemes be evaluated in real-world CRAHNs
{"title":"Optimising the Cuckoo Search Algorithm for Improved Quality of Service in Cognitive Radio ad hoc Networks","authors":"Ramahlapane Lerato Moila, M. Velempini","doi":"10.1109/icABCD59051.2023.10220569","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220569","url":null,"abstract":"This study proposes an optimised routing scheme, called OCS-AODV, for Cognitive Radio Ad Hoc Networks (CRAHNs) to enhance Quality of Service (QoS). The scheme applies the Cuckoo Search (CS) algorithm optimised with a fitness function to improve the performance of the Ad Hoc On-Demand Distance Vector (AODV). The objective of the study is to evaluate the proposed scheme's performance with respect to delay, packet loss, packet delivery ratio and throughput. The literature review shows that the existing routing protocols have limitations which impact performance in dynamic environments. The proposed OCS-AODV scheme aims to address these limitations by selecting reliable paths based on a fitness function that considers the lifetime of nodes, reliability, and available buffer capacity. The simulation results have shown that the OCS-AODV scheme outperforms the CS-DSDV and ACO-AODV schemes in terms of PDR, packet loss, delay, and throughput. The study concludes that the proposed scheme improves the QoS of routing in CRAHNs. However, the use of a single fitness function may not be optimal for all network scenarios. Multiple fitness functions may be considered in future and the schemes be evaluated in real-world CRAHNs","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"74 1","pages":"1-5"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74075792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220457
A. Periola, M. Sumbwanyambe
Ice melting in the Arctic enables the conduct of underwater neutrino astronomy in new regions with maritime resources. The presented research proposes a novel underwater network that is integrated with terrestrial computing entities to obtain underwater astronomy-associated data. In addition, the proposed network architecture enhances the conduct of underwater neutrino astronomy. This is done by increasing the potential neutrino presence points. Analysis shows that the use of the arctic region in addition to the existing region of Lake Baikal in comparison to the existing case (where only Lake Baikal is utilized) increases the potential neutrino presence points by an average of (28.3 – 65.7) %.
{"title":"An Underwater Network for Mini-Submarine Underwater Observatory","authors":"A. Periola, M. Sumbwanyambe","doi":"10.1109/icABCD59051.2023.10220457","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220457","url":null,"abstract":"Ice melting in the Arctic enables the conduct of underwater neutrino astronomy in new regions with maritime resources. The presented research proposes a novel underwater network that is integrated with terrestrial computing entities to obtain underwater astronomy-associated data. In addition, the proposed network architecture enhances the conduct of underwater neutrino astronomy. This is done by increasing the potential neutrino presence points. Analysis shows that the use of the arctic region in addition to the existing region of Lake Baikal in comparison to the existing case (where only Lake Baikal is utilized) increases the potential neutrino presence points by an average of (28.3 – 65.7) %.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"118 1","pages":"1-6"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75011344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220517
Malithi De Silva, Dane Brown
Enhancing agricultural practices has become essential in mitigating global hunger. Over the years, significant technological advancements have been introduced to improve the quality and quantity of harvests by effectively managing weeds, pests, and diseases. Many studies have focused on identifying plant diseases, as this information aids in making informed decisions about applying fungicides and fertilizers. Advanced systems often employ a combination of image processing and deep learning techniques to identify diseases based on visible symptoms. However, these systems typically rely on pre-existing datasets or images captured in controlled environments. This study showcases the efficacy of utilizing multispectral images captured in visible and Near Infrared (NIR) ranges for identifying plant diseases in real-world environmental conditions. The collected datasets were classified using popular Vision Transformer (ViT) models, including ViT- S16, ViT-BI6, ViT-LI6 and ViT-B32. The results showed impressive training and test accuracies for all the data collected using diverse Kolari vision lenses with 93.71 % and 90.02 %, respectively. This work highlights the potential of utilizing advanced imaging techniques for accurate and reliable plant disease identification in practical field conditions.
{"title":"Plant Disease Detection using Vision Transformers on Multispectral Natural Environment Images","authors":"Malithi De Silva, Dane Brown","doi":"10.1109/icABCD59051.2023.10220517","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220517","url":null,"abstract":"Enhancing agricultural practices has become essential in mitigating global hunger. Over the years, significant technological advancements have been introduced to improve the quality and quantity of harvests by effectively managing weeds, pests, and diseases. Many studies have focused on identifying plant diseases, as this information aids in making informed decisions about applying fungicides and fertilizers. Advanced systems often employ a combination of image processing and deep learning techniques to identify diseases based on visible symptoms. However, these systems typically rely on pre-existing datasets or images captured in controlled environments. This study showcases the efficacy of utilizing multispectral images captured in visible and Near Infrared (NIR) ranges for identifying plant diseases in real-world environmental conditions. The collected datasets were classified using popular Vision Transformer (ViT) models, including ViT- S16, ViT-BI6, ViT-LI6 and ViT-B32. The results showed impressive training and test accuracies for all the data collected using diverse Kolari vision lenses with 93.71 % and 90.02 %, respectively. This work highlights the potential of utilizing advanced imaging techniques for accurate and reliable plant disease identification in practical field conditions.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"5 1","pages":"1-6"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75417788","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-03DOI: 10.1109/icABCD59051.2023.10220521
Luxolo Kuhlane, Dane Brown, Marc Marais
This paper proposes a real-time system for detecting and tracking squids using the YOLOv5 object detection algorithm. The system utilizes a large dataset of annotated squid images and videos to train a YOLOv5 model optimized for detecting and tracking squids. The model is fine-tuned to minimize false positives and optimize detection accuracy. The system is deployed on a GPU-enabled device for real-time processing of video streams and tracking of detected squids across frames. The accuracy and speed of the system make it a valuable tool for marine scientists, conservationists, and fishermen to better understand the behavior and distribution of these elusive creatures. Future work includes incorporating additional computer vision techniques and sensor data to improve tracking accuracy and robustness.
{"title":"Real- Time Detecting and Tracking of Squids Using YOLOv5","authors":"Luxolo Kuhlane, Dane Brown, Marc Marais","doi":"10.1109/icABCD59051.2023.10220521","DOIUrl":"https://doi.org/10.1109/icABCD59051.2023.10220521","url":null,"abstract":"This paper proposes a real-time system for detecting and tracking squids using the YOLOv5 object detection algorithm. The system utilizes a large dataset of annotated squid images and videos to train a YOLOv5 model optimized for detecting and tracking squids. The model is fine-tuned to minimize false positives and optimize detection accuracy. The system is deployed on a GPU-enabled device for real-time processing of video streams and tracking of detected squids across frames. The accuracy and speed of the system make it a valuable tool for marine scientists, conservationists, and fishermen to better understand the behavior and distribution of these elusive creatures. Future work includes incorporating additional computer vision techniques and sensor data to improve tracking accuracy and robustness.","PeriodicalId":51314,"journal":{"name":"Big Data","volume":"17 1","pages":"1-5"},"PeriodicalIF":4.6,"publicationDate":"2023-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82658312","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","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：