Pub Date : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205761
Rio Junior, Ary Murti, Dien Rahmawati
An earthquake is one disaster that happened unpredictably and in some cases, it harms humanity. There are lots of research that studies earthquake vibrations using machine learning algorithms. However, implementing it in real-time application systems such as early warning systems is quite challenging due to the similarity of earthquake vibrations and non-earthquake vibrations (human activities and noises). Therefore, this study proposed an earthquake detection with Random Forest Classifier to distinguish earthquake and non-earthquake vibrations in a real-time application earthquake detection system. This study shows that Random Forest Classifier in a detection device is capable of classifying non-earthquake vibrations very well while it can classify earthquake vibrations with a success rate of 78.89%.
{"title":"Implementation of Random Forest Classifier for Real-time Earthquake Detection System","authors":"Rio Junior, Ary Murti, Dien Rahmawati","doi":"10.1109/IAICT59002.2023.10205761","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205761","url":null,"abstract":"An earthquake is one disaster that happened unpredictably and in some cases, it harms humanity. There are lots of research that studies earthquake vibrations using machine learning algorithms. However, implementing it in real-time application systems such as early warning systems is quite challenging due to the similarity of earthquake vibrations and non-earthquake vibrations (human activities and noises). Therefore, this study proposed an earthquake detection with Random Forest Classifier to distinguish earthquake and non-earthquake vibrations in a real-time application earthquake detection system. This study shows that Random Forest Classifier in a detection device is capable of classifying non-earthquake vibrations very well while it can classify earthquake vibrations with a success rate of 78.89%.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127423713","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205874
V. Vu, Y. Ivanenko, Thomas K. Sjögren, M. Pettersson
The sub-THz and THz frequency ranges, that have been designated for astronomy and military, are now considered for the next generation of cellular networks. Radar systems operating in these frequency ranges are popular. Sharing the same radio frequency (RF) resources of cellular networks and radar systems opens the great potential to integrate radar system on mobile equipment. With this integration, the radar applications based on detection and ranging are available for mobile equipment. Realizing synthetic aperture radar (SAR) is also possible due to the movement of the mobile equipment that helps to synthesize an aperture larger than the physical aperture of the mobile equipment. Precise active localization in indoor environment is therefore feasible. The paper presents a discussion about active localization by realizing monostatic, bistatic, multistatic and passive SAR on mobile equipment with an integrated radar system. The simulation and experiment results shows the feasibility of the proposal.
{"title":"Realizing SAR for Localization on Mobile Equipment with Integrated Radar System","authors":"V. Vu, Y. Ivanenko, Thomas K. Sjögren, M. Pettersson","doi":"10.1109/IAICT59002.2023.10205874","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205874","url":null,"abstract":"The sub-THz and THz frequency ranges, that have been designated for astronomy and military, are now considered for the next generation of cellular networks. Radar systems operating in these frequency ranges are popular. Sharing the same radio frequency (RF) resources of cellular networks and radar systems opens the great potential to integrate radar system on mobile equipment. With this integration, the radar applications based on detection and ranging are available for mobile equipment. Realizing synthetic aperture radar (SAR) is also possible due to the movement of the mobile equipment that helps to synthesize an aperture larger than the physical aperture of the mobile equipment. Precise active localization in indoor environment is therefore feasible. The paper presents a discussion about active localization by realizing monostatic, bistatic, multistatic and passive SAR on mobile equipment with an integrated radar system. The simulation and experiment results shows the feasibility of the proposal.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132731293","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205920
Hussain Salman, Eman Almohsen, M. Aljawder, A. Althawadi
The Kingdom of Bahrain has launched various mobile government applications that work side by side with the national e-government portal by providing a range of government services, while services offered on mobile applications are still limited compared to the e-government portal, utilization of end users’ feedback is vital to improve and enhance functionality to ensure proper digital integration on the mobile environment. In this research, knowledge engineering using natural language processing is implemented to analyze 20,000 user reviews of the top four most reviewed google play mobile government applications in Bahrain. Two resampling techniques were used to under-sample and over-sample unbalanced datasets; Near-Miss and Synthetic Minority Oversampling combined with Edited Nearest Neighbor. The performance of three classifiers for data analysis was compared and assessed before and after data resampling. Results suggest that the Random Forest classifier outperformed Artificial Neural Network and LogitBoost.
{"title":"Knowledge Engineering Using Natural Language Processing of User Reviews for Bahrain’s Mobile Government Applications","authors":"Hussain Salman, Eman Almohsen, M. Aljawder, A. Althawadi","doi":"10.1109/IAICT59002.2023.10205920","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205920","url":null,"abstract":"The Kingdom of Bahrain has launched various mobile government applications that work side by side with the national e-government portal by providing a range of government services, while services offered on mobile applications are still limited compared to the e-government portal, utilization of end users’ feedback is vital to improve and enhance functionality to ensure proper digital integration on the mobile environment. In this research, knowledge engineering using natural language processing is implemented to analyze 20,000 user reviews of the top four most reviewed google play mobile government applications in Bahrain. Two resampling techniques were used to under-sample and over-sample unbalanced datasets; Near-Miss and Synthetic Minority Oversampling combined with Edited Nearest Neighbor. The performance of three classifiers for data analysis was compared and assessed before and after data resampling. Results suggest that the Random Forest classifier outperformed Artificial Neural Network and LogitBoost.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"338 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132419246","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205894
A. Boumaiza, A. Sanfilippo
The use of distributed energy generation through business and residential photovoltaic (PV) applications creates new energy markets that blur the traditional line between energy providers and users. This new market dynamic results in the emergence of energy prosumers, whose role is to produce and consume energy. Blockchain technology automates direct energy exchanges within a distributed system architecture that relies on encryption hashing and general agreement verification. This technology provides prosumers, consumers, energy providers, and utilities with an affordable, safe, and unique energy-trading alternative. The Education City Community Housing (ECCH) in Qatar is the focus of this project, which aims to develop and implement an accurate Agent-Based Modeling (ABM) model and a Geographic Information System (GIS) to facilitate energy exchange in a real estate market. The ABM model simulates the spatiotemporal aspects of trading in a small market and collects and analyzes a large amount of data about daily energy usage. These simulations can help to better understand the structure of a trading market and to develop a decentralized system for trading energy. The findings of this study demonstrate that the peculiarities of transactions carried out in a community-based housing market can be easily researched using GIS data combined with an agent-based design by simply changing the settings. For large-scale simulation models with numerous stakeholders, high-performance computing will be used to improve the model’s performance and to provide a scalable environment for analyzing an energy blockchain community for the technological, financial, and social sectors of Qatar.
{"title":"Blockchain-based Electricity Market Agent-based Modelling&Simulation","authors":"A. Boumaiza, A. Sanfilippo","doi":"10.1109/IAICT59002.2023.10205894","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205894","url":null,"abstract":"The use of distributed energy generation through business and residential photovoltaic (PV) applications creates new energy markets that blur the traditional line between energy providers and users. This new market dynamic results in the emergence of energy prosumers, whose role is to produce and consume energy. Blockchain technology automates direct energy exchanges within a distributed system architecture that relies on encryption hashing and general agreement verification. This technology provides prosumers, consumers, energy providers, and utilities with an affordable, safe, and unique energy-trading alternative. The Education City Community Housing (ECCH) in Qatar is the focus of this project, which aims to develop and implement an accurate Agent-Based Modeling (ABM) model and a Geographic Information System (GIS) to facilitate energy exchange in a real estate market. The ABM model simulates the spatiotemporal aspects of trading in a small market and collects and analyzes a large amount of data about daily energy usage. These simulations can help to better understand the structure of a trading market and to develop a decentralized system for trading energy. The findings of this study demonstrate that the peculiarities of transactions carried out in a community-based housing market can be easily researched using GIS data combined with an agent-based design by simply changing the settings. For large-scale simulation models with numerous stakeholders, high-performance computing will be used to improve the model’s performance and to provide a scalable environment for analyzing an energy blockchain community for the technological, financial, and social sectors of Qatar.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"3 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133174324","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205588
Abd Salam At Taqwa, Z. Zainuddin, Z. Tahir
3D Morphable Model, one of the models used to reconstruct 3D face from 2D monocular image of face, has achieved satisfactory results along with computer vision and graphics development. However, reconstructing 3D face using a 3D Morphable Model in a weakly-supervised manner has its challenges because it does not require labels as ground truth and only relies on the similarity of features between 2D monocular image and 3D face. This research uses weakly-supervised 3D face reconstruction by comparing identity feature extraction. In this case, deep face recognition techniques used for identity feature extraction are ArcFace, CosFace, and ElasticFace. The 3D face reconstruction process is divided into 1) rigid fitting to fit the 3D face landmarks into face landmarks of 2D monocular image and 2) non-rigid fitting feature similarity with hybrid-level weak supervision applying diverse deep face recognition models. The results of the reconstruction are subsequently evaluated using the NoW challenge. Experimental results on the NoW protocol show that ElasticFace-Arc is the best deep face recognition for monocular 3d face reconstruction.
{"title":"Monocular 3D Face Reconstruction Using 3D Morphable Model and ElasticFace","authors":"Abd Salam At Taqwa, Z. Zainuddin, Z. Tahir","doi":"10.1109/IAICT59002.2023.10205588","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205588","url":null,"abstract":"3D Morphable Model, one of the models used to reconstruct 3D face from 2D monocular image of face, has achieved satisfactory results along with computer vision and graphics development. However, reconstructing 3D face using a 3D Morphable Model in a weakly-supervised manner has its challenges because it does not require labels as ground truth and only relies on the similarity of features between 2D monocular image and 3D face. This research uses weakly-supervised 3D face reconstruction by comparing identity feature extraction. In this case, deep face recognition techniques used for identity feature extraction are ArcFace, CosFace, and ElasticFace. The 3D face reconstruction process is divided into 1) rigid fitting to fit the 3D face landmarks into face landmarks of 2D monocular image and 2) non-rigid fitting feature similarity with hybrid-level weak supervision applying diverse deep face recognition models. The results of the reconstruction are subsequently evaluated using the NoW challenge. Experimental results on the NoW protocol show that ElasticFace-Arc is the best deep face recognition for monocular 3d face reconstruction.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124455687","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205661
N. Shabrina, Ryukin Aranta Lika, S. Indarti
Plant-parasitic nematodes are major agricultural pathogens contributing to massive crop losses worldwide. It is crucial to identify plant-parasitic nematodes to decide the best pest control and management strategy. The current identification technique is based on visual observation from nematode microscopic images done by the nematologist. However, this method requires a long process and is prone to error. A deep learning-based method can be implemented to speed up the current identification process. This study explores the effect of combining several data augmentation techniques, namely brightness, contrast, blur, and noise, on the performance of the EfficientNetV2B0 and EfficientNetV2M models for identifying plant-parasitic nematodes. Moreover, this study also compared three optimizers while training the models to find the best optimizer for each model and data augmentation. The results show that the EfficientNetV2B0 model yielded the highest test accuracy of 96.91% when employing no augmentation and trained using SGD and RMSProp optimizer. Furthermore, the EfficientNetV2M model gave the highest test accuracy of 96.91% when the combination of brightness and contrast augmentations was applied and trained using the RMSProp optimizer.
{"title":"The Effect of Data Augmentation and Optimization Technique on the Performance of EfficientNetV2 for Plant-Parasitic Nematode Identification","authors":"N. Shabrina, Ryukin Aranta Lika, S. Indarti","doi":"10.1109/IAICT59002.2023.10205661","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205661","url":null,"abstract":"Plant-parasitic nematodes are major agricultural pathogens contributing to massive crop losses worldwide. It is crucial to identify plant-parasitic nematodes to decide the best pest control and management strategy. The current identification technique is based on visual observation from nematode microscopic images done by the nematologist. However, this method requires a long process and is prone to error. A deep learning-based method can be implemented to speed up the current identification process. This study explores the effect of combining several data augmentation techniques, namely brightness, contrast, blur, and noise, on the performance of the EfficientNetV2B0 and EfficientNetV2M models for identifying plant-parasitic nematodes. Moreover, this study also compared three optimizers while training the models to find the best optimizer for each model and data augmentation. The results show that the EfficientNetV2B0 model yielded the highest test accuracy of 96.91% when employing no augmentation and trained using SGD and RMSProp optimizer. Furthermore, the EfficientNetV2M model gave the highest test accuracy of 96.91% when the combination of brightness and contrast augmentations was applied and trained using the RMSProp optimizer.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125532795","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205795
R. Kurniawan, I. Iskandar, F. Lestari, Habibi Al Rasyid Harpizon, Ilyas Husti
YouTube is a widely-used platform in Indonesia, with 93.8% of its users. As such, it presents a valuable opportunity for marketing tourist destinations, particularly in Riau province, which aims to become Indonesia’s top Halal travel destination. Tourism is a vital contributor to the economic growth of regions, and each province in Indonesia competes to promote its tourist attractions to attract more visitors every year. However, the large volume of data can challenge the manual analysis of feedback from YouTube’s features, such as likes, dislikes, and comments. A literature review suggests that the Naive Bayes algorithm, which uses machine learning, is helpful for sentiment analysis. Therefore, this study aims to analyze public sentiment toward tourist destinations in Riau province by analyzing YouTube comments using the Naïve Bayes algorithm. The study used 1680 opinions collected from 10 YouTube videos showcasing tourist destinations in Riau. The Naive Bayes algorithm classified 60% of the comments as positive, 32% as neutral, and 8% as negative. The experimental results indicated an accuracy and precision of 73%, a recall of 94%, and an F-1 Score of 82%. The study used the word frequency technique to reveal that Riau could become a popular halal tourist destination based on several frequently occurring words in the comments.
{"title":"Exploring Tourist Feedback on Riau Attractions Through Indonesian Language YouTube Opinion Using Naïve Bayes Algorithm","authors":"R. Kurniawan, I. Iskandar, F. Lestari, Habibi Al Rasyid Harpizon, Ilyas Husti","doi":"10.1109/IAICT59002.2023.10205795","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205795","url":null,"abstract":"YouTube is a widely-used platform in Indonesia, with 93.8% of its users. As such, it presents a valuable opportunity for marketing tourist destinations, particularly in Riau province, which aims to become Indonesia’s top Halal travel destination. Tourism is a vital contributor to the economic growth of regions, and each province in Indonesia competes to promote its tourist attractions to attract more visitors every year. However, the large volume of data can challenge the manual analysis of feedback from YouTube’s features, such as likes, dislikes, and comments. A literature review suggests that the Naive Bayes algorithm, which uses machine learning, is helpful for sentiment analysis. Therefore, this study aims to analyze public sentiment toward tourist destinations in Riau province by analyzing YouTube comments using the Naïve Bayes algorithm. The study used 1680 opinions collected from 10 YouTube videos showcasing tourist destinations in Riau. The Naive Bayes algorithm classified 60% of the comments as positive, 32% as neutral, and 8% as negative. The experimental results indicated an accuracy and precision of 73%, a recall of 94%, and an F-1 Score of 82%. The study used the word frequency technique to reveal that Riau could become a popular halal tourist destination based on several frequently occurring words in the comments.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"111 3S 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126985146","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205825
A. Khurana, Vv Ravikumar, Vinod Kumar
The increasing application of Artificial Intelligence usage in health care and medicine has attracted considerable research interest in the recent past. The objective of this paper was to provide a scientometric analysis of artificial intelligence in the healthcare sector. The momentum gained by artificial intelligence and information technology gave impetus and importance to health care. The researchers used the Scopus database to extract the papers. VosViewer tool was employed for advanced analysis, whereas Microsoft Excel was used in depicting graphical representation. The study summarises the research conducted by different authors, various universities and countries to extend the benefits of artificial intelligence to healthcare. This paper would be of some help to practitioners and researchers in the healthcare sector to know more about AI in healthcare.
{"title":"A Scientometric Study of Artificial Intelligence in the Health Care Sector","authors":"A. Khurana, Vv Ravikumar, Vinod Kumar","doi":"10.1109/IAICT59002.2023.10205825","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205825","url":null,"abstract":"The increasing application of Artificial Intelligence usage in health care and medicine has attracted considerable research interest in the recent past. The objective of this paper was to provide a scientometric analysis of artificial intelligence in the healthcare sector. The momentum gained by artificial intelligence and information technology gave impetus and importance to health care. The researchers used the Scopus database to extract the papers. VosViewer tool was employed for advanced analysis, whereas Microsoft Excel was used in depicting graphical representation. The study summarises the research conducted by different authors, various universities and countries to extend the benefits of artificial intelligence to healthcare. This paper would be of some help to practitioners and researchers in the healthcare sector to know more about AI in healthcare.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"133 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117245519","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205657
Amos Noel, Wougens Vincent, J. Piou
In this paper, the residual convolutional neural network Resnet50 is applied to satellite imagery collected on the 12 January 2010 earthquake with a moment magnitude (Mw) of 7.0 that struck the western cities of Haiti such as Port-au-Prince, Leogane and Jacmel, and on 14 August 2021 another earthquake with moment magnitude (Mw) of 7.2 that hit the southwestern peninsula of Haiti with its epicenter located not too far from the main city of Les Cayes. Meta data that provide geolocations of landmark buildings, residential quarters, road infrastructures and landslide areas are used to partition the post-earthquake satellite images and create three class databanks that allow training and testing of the Resnet50 architecture to establish similarities between western and southwestern areas of the country in land topography, housing quarters and road networks. In a first experiment, datasets derived from the post-earthquake image of 12 January 2010 are used to train the network while the datasets from the post-earthquake of 14 August 2021 are reserved for testing; the network architecture Resnet50 exhibits an average performance of about 88% on testing. Using data augmentation by 8 fold on the training set with datasets from the 14 August 2021 earthquake, testing performance on the 12 January 2010 earthquake improves by 4% with the network trained on the original datasets. Therefore, Resnet50 appears to be a well suited network architecture to detect and locate land areas, houses and roads severely impacted by an earthquake.
{"title":"Resnet50 to Detect Landslides, Damaged Infrastructures and Crumbled Houses from Haiti 2010 and 2021 Earthquakes","authors":"Amos Noel, Wougens Vincent, J. Piou","doi":"10.1109/IAICT59002.2023.10205657","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205657","url":null,"abstract":"In this paper, the residual convolutional neural network Resnet50 is applied to satellite imagery collected on the 12 January 2010 earthquake with a moment magnitude (Mw) of 7.0 that struck the western cities of Haiti such as Port-au-Prince, Leogane and Jacmel, and on 14 August 2021 another earthquake with moment magnitude (Mw) of 7.2 that hit the southwestern peninsula of Haiti with its epicenter located not too far from the main city of Les Cayes. Meta data that provide geolocations of landmark buildings, residential quarters, road infrastructures and landslide areas are used to partition the post-earthquake satellite images and create three class databanks that allow training and testing of the Resnet50 architecture to establish similarities between western and southwestern areas of the country in land topography, housing quarters and road networks. In a first experiment, datasets derived from the post-earthquake image of 12 January 2010 are used to train the network while the datasets from the post-earthquake of 14 August 2021 are reserved for testing; the network architecture Resnet50 exhibits an average performance of about 88% on testing. Using data augmentation by 8 fold on the training set with datasets from the 14 August 2021 earthquake, testing performance on the 12 January 2010 earthquake improves by 4% with the network trained on the original datasets. Therefore, Resnet50 appears to be a well suited network architecture to detect and locate land areas, houses and roads severely impacted by an earthquake.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"254 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117324589","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 : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205595
Mikho J Pelingon, Valenzuela Franco Carlos, M. L. Guico, J. K. Galicia
Calamansi has been declared as one of the most important fruit growing crops in the Philippines. However, due to certain bacteria, it is susceptible to certain diseases affecting its harvest rate. This paper aims to effectively monitor the state of the calamansi at its healthy state and at its diseased state. Specifically, it classifies diseases such as Citrus Canker, Citrus Scab, and Citrus Browning by utilizing existing image processing techniques for disease detection of different fruits and determining which algorithm is most apt for this application in terms of precision, accuracy and recall. Techniques such as K-Means Clustering, utilization of an Artificial Neural Network (ANN), feature extraction through GLCM along with the usage of a minimum distance classifier, a Support Vector Machine (SVM) classifier and other techniques and/or their combinations were explored and measured. The researchers performed two kinds of tests: 1×1 comparison and merged comparison. For the 1×1 comparison, making use of GrabCut, color feature extraction, and SVM produced the best overall results, with an overall average of 98% for precision, 95% for accuracy, 91% for recall, and 94% for F-score. Adaptive Gaussian Filtering along with texture feature extraction and SVM was the most accurate for detecting calamansi fruits with citrus canker and citrus scab. Overall, the two methods acquired the same average accuracy of 61%
{"title":"Application of Image Recognition Algorithms in the Detection of Philippine Lime Diseases","authors":"Mikho J Pelingon, Valenzuela Franco Carlos, M. L. Guico, J. K. Galicia","doi":"10.1109/IAICT59002.2023.10205595","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205595","url":null,"abstract":"Calamansi has been declared as one of the most important fruit growing crops in the Philippines. However, due to certain bacteria, it is susceptible to certain diseases affecting its harvest rate. This paper aims to effectively monitor the state of the calamansi at its healthy state and at its diseased state. Specifically, it classifies diseases such as Citrus Canker, Citrus Scab, and Citrus Browning by utilizing existing image processing techniques for disease detection of different fruits and determining which algorithm is most apt for this application in terms of precision, accuracy and recall. Techniques such as K-Means Clustering, utilization of an Artificial Neural Network (ANN), feature extraction through GLCM along with the usage of a minimum distance classifier, a Support Vector Machine (SVM) classifier and other techniques and/or their combinations were explored and measured. The researchers performed two kinds of tests: 1×1 comparison and merged comparison. For the 1×1 comparison, making use of GrabCut, color feature extraction, and SVM produced the best overall results, with an overall average of 98% for precision, 95% for accuracy, 91% for recall, and 94% for F-score. Adaptive Gaussian Filtering along with texture feature extraction and SVM was the most accurate for detecting calamansi fruits with citrus canker and citrus scab. Overall, the two methods acquired the same average accuracy of 61%","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127833633","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：