Pub Date : 2023-07-13DOI: 10.1109/IAICT59002.2023.10205726
H. Dutta, K. Manivas, Marjana Bhuyan, M. Bhuyan
Hand gesture recognition is one of the interesting problems of Computer Vision. It has a wide range of applications in the fields of Human-Computer Interaction, Robotics, Sign language interpretation, Augmented Reality, etc. Most of the existing deep learning methods detect hand gestures in two stages. The hand is located in the first stage, and classification is performed on the hand portion in the second stage to estimate the hand pose. Although these methods are accurate, they are slow and cant be used for real-time applications. Few existing literature even explored one-stage approaches, like YOLO, SSD, etc., for hand gesture recognition as they have less inference time. But they place many anchor boxes over an image of which only a small percentage are positive. This leads to a huge imbalance between positive and negative anchor boxes and slows the training process. In this paper, we have used an end-to-end, one-stage hand detection-based approach, namely, CenterNet, for hand gesture recognition. It detects the object as a point, i.e., the center point of the bounding box encompassing the object, and regresses to the object size. This eliminates the need for anchor boxes in CenterNet. We have added Dual Attention Network to the CenterNet architecture to improve the performance. Our model achieves a mean F1-score of 84.40% and 98.83% on Ouhands and NUS hand pose datasets, respectively. Results show that our model can perform well even under complex backgrounds and varying illumination conditions, and the F1-scores obtained are close to benchmark values.
{"title":"An End-to-end Anchorless Approach to Recognize Hand Gestures using CenterNet","authors":"H. Dutta, K. Manivas, Marjana Bhuyan, M. Bhuyan","doi":"10.1109/IAICT59002.2023.10205726","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205726","url":null,"abstract":"Hand gesture recognition is one of the interesting problems of Computer Vision. It has a wide range of applications in the fields of Human-Computer Interaction, Robotics, Sign language interpretation, Augmented Reality, etc. Most of the existing deep learning methods detect hand gestures in two stages. The hand is located in the first stage, and classification is performed on the hand portion in the second stage to estimate the hand pose. Although these methods are accurate, they are slow and cant be used for real-time applications. Few existing literature even explored one-stage approaches, like YOLO, SSD, etc., for hand gesture recognition as they have less inference time. But they place many anchor boxes over an image of which only a small percentage are positive. This leads to a huge imbalance between positive and negative anchor boxes and slows the training process. In this paper, we have used an end-to-end, one-stage hand detection-based approach, namely, CenterNet, for hand gesture recognition. It detects the object as a point, i.e., the center point of the bounding box encompassing the object, and regresses to the object size. This eliminates the need for anchor boxes in CenterNet. We have added Dual Attention Network to the CenterNet architecture to improve the performance. Our model achieves a mean F1-score of 84.40% and 98.83% on Ouhands and NUS hand pose datasets, respectively. Results show that our model can perform well even under complex backgrounds and varying illumination conditions, and the F1-scores obtained are close to benchmark values.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"12 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":"114379220","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/IAICT59002.2023.10205599
Laode Muh, AM Armadi, Indrabayu, I. Nurtanio
This research aims to detect and classify snacks. The detection and classification process uses the Mask R-CNN algorithm. The training process is carried out in the training stage with 250 epochs and 150 steps per epoch. The dataset used in this study consists of 687 snack images with a resolution of 640 x640 pixels divided into 549 training data and 137 validation data. In addition, System testing results were conducted using scenarios 1-7 in an overlapping or partially covered state within the 10-70% range. It can be interpreted that snack overlap detection has optimal performance in the 10-50% range, as evidenced by the high mAP value of 0.99. However, the system cannot detect well in the 60% and 70% overlap range, as seen from the low mAP values of only 0.2 and 0. The evaluation results show that the system has an excellent performance in performing object detection and classification tasks with high accuracy and consistency.
{"title":"Snacks Detection Under Overlapped Conditions Using Computer Vision","authors":"Laode Muh, AM Armadi, Indrabayu, I. Nurtanio","doi":"10.1109/IAICT59002.2023.10205599","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205599","url":null,"abstract":"This research aims to detect and classify snacks. The detection and classification process uses the Mask R-CNN algorithm. The training process is carried out in the training stage with 250 epochs and 150 steps per epoch. The dataset used in this study consists of 687 snack images with a resolution of 640 x640 pixels divided into 549 training data and 137 validation data. In addition, System testing results were conducted using scenarios 1-7 in an overlapping or partially covered state within the 10-70% range. It can be interpreted that snack overlap detection has optimal performance in the 10-50% range, as evidenced by the high mAP value of 0.99. However, the system cannot detect well in the 60% and 70% overlap range, as seen from the low mAP values of only 0.2 and 0. The evaluation results show that the system has an excellent performance in performing object detection and classification tasks with high accuracy and consistency.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"91 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":"114717008","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/IAICT59002.2023.10205909
A. Suharjono, M. Mukhlisin, E. Wardihani, Muhlasah Novitasari, Efrilia M Khusna, Dara Aulia Feryando, W. Adi, S. Pramono, R. Apriantoro, Irfan Mujahidin
Propagation along railway crossings exhibits different propagation characteristics compared to the general environment, especially in Indonesia. The environmental conditions along the railway crossings vary, including straight tracks, track turns, track bridges, and tunnel tracks, resulting in pathloss that originates from the material structure of the railroad tracks. The purpose of this research is to determine the path loss coefficient values (n) under railway crossing conditions using LoRa system transmission performance with a sample area in Indonesia. The method involves measuring the RSSI (Signal Strength parameter indicator) values of LoRa nodes under various environmental conditions, including straight tracks, track turns, track bridges, and tunnel tracks. Based on the results and analysis, the value of n for RSSI under straight track conditions with no railway crossing was found to be 1.948, while it increased to 2.4929 when trains passed through. Under turn track conditions with no trains passing, the value of n was found to be 1.8646.
{"title":"Performance Evaluation of LoRa 915 MHz for IoT Communication System on Indonesian Railway Tracks with Environmental Factor Propagation Analysis","authors":"A. Suharjono, M. Mukhlisin, E. Wardihani, Muhlasah Novitasari, Efrilia M Khusna, Dara Aulia Feryando, W. Adi, S. Pramono, R. Apriantoro, Irfan Mujahidin","doi":"10.1109/IAICT59002.2023.10205909","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205909","url":null,"abstract":"Propagation along railway crossings exhibits different propagation characteristics compared to the general environment, especially in Indonesia. The environmental conditions along the railway crossings vary, including straight tracks, track turns, track bridges, and tunnel tracks, resulting in pathloss that originates from the material structure of the railroad tracks. The purpose of this research is to determine the path loss coefficient values (n) under railway crossing conditions using LoRa system transmission performance with a sample area in Indonesia. The method involves measuring the RSSI (Signal Strength parameter indicator) values of LoRa nodes under various environmental conditions, including straight tracks, track turns, track bridges, and tunnel tracks. Based on the results and analysis, the value of n for RSSI under straight track conditions with no railway crossing was found to be 1.948, while it increased to 2.4929 when trains passed through. Under turn track conditions with no trains passing, the value of n was found to be 1.8646.","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":"124127957","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/IAICT59002.2023.10205864
Made Adi Paramartha Putra, G. Sampedro, Dong‐Seong Kim, Jae-Min Lee
This research paper proposes a multi-criteria client selection approach to enhance the efficiency of Federated Learning (FL). While the state-of-the-art client selection in FL mainly focuses on a single characteristic to determine a suitable client for the training process, a multi-criteria selection is needed to provide a more efficient FL system. We introduce the Ensembled Client Selection Mechanism (ECSM) as a novel approach to address this issue. The proposed approach takes into account client accuracy, reputation, and randomness to improve accuracy during the lower communication period. The study employs random client selection to prevent repetitive training and ensure model generalization. The results indicate that the proposed ECSM mechanism can improve FL performance by achieving the desired accuracy with fewer communication rounds. Specifically, the approach improves FL efficiency by 56% when tested on the FMNIST dataset compared to the baseline approach. These findings suggest that the ECSM mechanism can significantly enhance the efficiency of the FL process.
{"title":"ECSM: An Ensembled Client Selection Mechanism for Efficient Federated Learning","authors":"Made Adi Paramartha Putra, G. Sampedro, Dong‐Seong Kim, Jae-Min Lee","doi":"10.1109/IAICT59002.2023.10205864","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205864","url":null,"abstract":"This research paper proposes a multi-criteria client selection approach to enhance the efficiency of Federated Learning (FL). While the state-of-the-art client selection in FL mainly focuses on a single characteristic to determine a suitable client for the training process, a multi-criteria selection is needed to provide a more efficient FL system. We introduce the Ensembled Client Selection Mechanism (ECSM) as a novel approach to address this issue. The proposed approach takes into account client accuracy, reputation, and randomness to improve accuracy during the lower communication period. The study employs random client selection to prevent repetitive training and ensure model generalization. The results indicate that the proposed ECSM mechanism can improve FL performance by achieving the desired accuracy with fewer communication rounds. Specifically, the approach improves FL efficiency by 56% when tested on the FMNIST dataset compared to the baseline approach. These findings suggest that the ECSM mechanism can significantly enhance the efficiency of the FL process.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"8 6 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":"131696330","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/IAICT59002.2023.10205853
Devaguptam Sreegeethi, Kogatam Thanmai, Lakshmi S Raj, D. Naik, Ranjit P. Kolkar
Most existing video stabilization techniques are used for post-processing, where previously recorded videos are given to the model to obtain stabilized versions. Online video stabilization usually relies on sensors like gyroscopes or assumes constant motion, which is not suitable for videos with changing motions. This work introduces a video stabilization technique with just one-frame latency. The algorithm operates at the spatial level in the infrequent domain, tracking the motion of mesh vertices. Motion tracks of feature marks are combined with the nearest mesh vertex using two median gauges, assigning each vertex a smooth motion track. The proposed approach, called anticipated foster track leveling, smoothes the motion profiles by utilizing previous motions and adapting accordingly for smoother results. This method can handle changes in movement in space and time and works in real-time, allowing applications in security systems, robotics, and unmanned aerial vehicles (UAVs). When evaluated against other models, MeshFlow gives an overall good performance in all comparison metrics evaluated. Hence MeshFlow can be used as a reliable low-latency technique for real-time video stabilization in remote devices.
{"title":"Online Video Stabilization using Mesh Flow with Minimum Latency","authors":"Devaguptam Sreegeethi, Kogatam Thanmai, Lakshmi S Raj, D. Naik, Ranjit P. Kolkar","doi":"10.1109/IAICT59002.2023.10205853","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205853","url":null,"abstract":"Most existing video stabilization techniques are used for post-processing, where previously recorded videos are given to the model to obtain stabilized versions. Online video stabilization usually relies on sensors like gyroscopes or assumes constant motion, which is not suitable for videos with changing motions. This work introduces a video stabilization technique with just one-frame latency. The algorithm operates at the spatial level in the infrequent domain, tracking the motion of mesh vertices. Motion tracks of feature marks are combined with the nearest mesh vertex using two median gauges, assigning each vertex a smooth motion track. The proposed approach, called anticipated foster track leveling, smoothes the motion profiles by utilizing previous motions and adapting accordingly for smoother results. This method can handle changes in movement in space and time and works in real-time, allowing applications in security systems, robotics, and unmanned aerial vehicles (UAVs). When evaluated against other models, MeshFlow gives an overall good performance in all comparison metrics evaluated. Hence MeshFlow can be used as a reliable low-latency technique for real-time video stabilization in remote devices.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"39 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":"123468210","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/IAICT59002.2023.10205951
Akhmad Sarif, D. Gunawan
Facial Expression Recognition (FER) through digital images has undergone significant development in line with the development of computer vision technology and artificial intelligence. Facial expression recognition that has utilized deep learning shows promising results. By using deep learning, classifying millions of digital images can be easier and more accurate. However, misclassification of facial expressions sometimes still occurs. This paper proposes a method for improving the AlexNet model for application in the FER area. Some pre-processing procedures were performed on the image dataset, including resizing the image size to 227x227, converting the image to RGB (Red Blue Green) format, adjusting the contrast level of the image using CLAHE (Contrast Limited Adaptive Histogram Equalization), and augmenting by cropping the dataset image. Meanwhile, fine-tuning the AlexNet model was done by changing the ReLU activation function to Leaky ReLU, input normalization from cross channel to batch normalization, and two dropout values (from 0.5 to 0.3 and 0), and changing the number of output classifications from 1000 to 7. The experimental results show that the proposed method enhances standard AlexNet’s performance by improving its accuracy to 24.82% on the CK+ dataset and 20.05% on the KDEF dataset. There is no misclassification of facial expressions when using the proposed method, as it occurs when using the standard AlexNet model.
{"title":"A Method for Improving AlexNet’s Performance in The Area of Facial Expressions Recognition","authors":"Akhmad Sarif, D. Gunawan","doi":"10.1109/IAICT59002.2023.10205951","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205951","url":null,"abstract":"Facial Expression Recognition (FER) through digital images has undergone significant development in line with the development of computer vision technology and artificial intelligence. Facial expression recognition that has utilized deep learning shows promising results. By using deep learning, classifying millions of digital images can be easier and more accurate. However, misclassification of facial expressions sometimes still occurs. This paper proposes a method for improving the AlexNet model for application in the FER area. Some pre-processing procedures were performed on the image dataset, including resizing the image size to 227x227, converting the image to RGB (Red Blue Green) format, adjusting the contrast level of the image using CLAHE (Contrast Limited Adaptive Histogram Equalization), and augmenting by cropping the dataset image. Meanwhile, fine-tuning the AlexNet model was done by changing the ReLU activation function to Leaky ReLU, input normalization from cross channel to batch normalization, and two dropout values (from 0.5 to 0.3 and 0), and changing the number of output classifications from 1000 to 7. The experimental results show that the proposed method enhances standard AlexNet’s performance by improving its accuracy to 24.82% on the CK+ dataset and 20.05% on the KDEF dataset. There is no misclassification of facial expressions when using the proposed method, as it occurs when using the standard AlexNet model.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"176 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":"123509878","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/IAICT59002.2023.10205837
R. Wibowo, Istikmal, A. Irawan
Internet of Things (IoT) is a network that connects various integrated objects. One application of IoT is a fire detection system to provide remote warnings. In this study, IoT deployments were performed using SVM (Support Vector Machine) algorithm and KNN (K-Nearest Neighbor) algorithm. The algorithm is attached to the ESP32 microcontroller for data classification. The sensors used include temperature, humidity, fire, and smoke sensors. In case of fire a warning will be sent to Telegram. Classification results were tested with Quality of Service (QoS) parameters on throughput, delay, and jitter values, as well as with the confusion matrix with 3 simulation variations. The test outcomes display that the system is in the correct category with an average throughput value of 1.848 bps and the best value of 1.858 bps, an average delay of 593.045 ms, and a jitter of 594.188 ms. The highest accuracy was obtained in simulation 2, namely 100% for SVM and 97.5% for KNN with K=1 in KNN. Meanwhile, in simulation 1 KNN has an accuracy of 95% and SVM 98%, simulation 3 KNN 97% and SVM 100%. Thus, the SVM algorithm can classify the system better than the KNN algorithm.
物联网(Internet of Things, IoT)是连接各种集成对象的网络。物联网的一个应用是提供远程警报的火灾探测系统。在本研究中，物联网部署使用SVM(支持向量机)算法和KNN (k -最近邻)算法进行。该算法附加在ESP32单片机上进行数据分类。传感器包括温度传感器、湿度传感器、火灾传感器和烟雾传感器。如果发生火灾，将向电报发送警告。使用吞吐量、延迟和抖动值的服务质量(QoS)参数以及具有3个模拟变量的混淆矩阵对分类结果进行测试。测试结果表明，系统处于正确的类别，平均吞吐量为1.848 bps，最佳值为1.858 bps，平均延迟为593.045 ms，抖动为594.188 ms。仿真2的准确率最高，SVM的准确率为100%，KNN中K=1的KNN准确率为97.5%。同时，仿真1中KNN的准确率为95%，SVM为98%，仿真3中KNN的准确率为97%，SVM为100%。因此，SVM算法可以比KNN算法更好地对系统进行分类。
{"title":"Comparison Analysis of SVM and KNN Algorithm For IoT-Based Home Fire Detection System","authors":"R. Wibowo, Istikmal, A. Irawan","doi":"10.1109/IAICT59002.2023.10205837","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205837","url":null,"abstract":"Internet of Things (IoT) is a network that connects various integrated objects. One application of IoT is a fire detection system to provide remote warnings. In this study, IoT deployments were performed using SVM (Support Vector Machine) algorithm and KNN (K-Nearest Neighbor) algorithm. The algorithm is attached to the ESP32 microcontroller for data classification. The sensors used include temperature, humidity, fire, and smoke sensors. In case of fire a warning will be sent to Telegram. Classification results were tested with Quality of Service (QoS) parameters on throughput, delay, and jitter values, as well as with the confusion matrix with 3 simulation variations. The test outcomes display that the system is in the correct category with an average throughput value of 1.848 bps and the best value of 1.858 bps, an average delay of 593.045 ms, and a jitter of 594.188 ms. The highest accuracy was obtained in simulation 2, namely 100% for SVM and 97.5% for KNN with K=1 in KNN. Meanwhile, in simulation 1 KNN has an accuracy of 95% and SVM 98%, simulation 3 KNN 97% and SVM 100%. Thus, the SVM algorithm can classify the system better than the KNN algorithm.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"178 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":"124417650","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/IAICT59002.2023.10205610
Rahmat, Z. Zainuddin, A. Achmad
Thermal imaging is a technology that utilizes heat radiation from objects, including duck eggs and has been widely used in the field of thermography. This study discusses a classification system of eggs using thermal camera images to differentiate between fertile and infertile eggs. The image processing methods used in this study are the histogram analysis and ROI method to identify the thermal characteristics of different eggs. The results of this study show that this method can distinguish between fertile and infertile eggs with high accuracy. This study can help farmers improve the efficiency of chicken reproduction and produce better-quality eggs. Therefore, this article has the potential to provide benefits for the livestock and food industries. In egg image processing, the ROI method increases analysis accuracy and classifies objects in the image. The histogram analysis method is used to provide accurate information. Testing with single and group egg images resulted in 93.7% accuracy in determining fertile and infertile eggs on the 9th day of incubation.
{"title":"Classification Of Fertile And Infertile Eggs Using Thermal Camera Image And Histogram Analysis: Technology Application In Poultry Farming Industry","authors":"Rahmat, Z. Zainuddin, A. Achmad","doi":"10.1109/IAICT59002.2023.10205610","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205610","url":null,"abstract":"Thermal imaging is a technology that utilizes heat radiation from objects, including duck eggs and has been widely used in the field of thermography. This study discusses a classification system of eggs using thermal camera images to differentiate between fertile and infertile eggs. The image processing methods used in this study are the histogram analysis and ROI method to identify the thermal characteristics of different eggs. The results of this study show that this method can distinguish between fertile and infertile eggs with high accuracy. This study can help farmers improve the efficiency of chicken reproduction and produce better-quality eggs. Therefore, this article has the potential to provide benefits for the livestock and food industries. In egg image processing, the ROI method increases analysis accuracy and classifies objects in the image. The histogram analysis method is used to provide accurate information. Testing with single and group egg images resulted in 93.7% accuracy in determining fertile and infertile eggs on the 9th day of incubation.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"25 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":"124668496","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/IAICT59002.2023.10205543
I. M. A. Wiryawan, D. D. Ariananda, S. Wibowo
In cognitive radio (CR) networks, secondary users (SUs) might be required to gauge a wide frequency band to find frequency holes that they can use for signal transmission. When this spectrum sensing process is conducted digitally, a high sampling rate might be needed to satisfy the Nyquist rate. However, the existence of the frequency holes can be concluded by simply constructing the power spectral density (PSD) instead of the original signal. In fact, the Nyquist criterion is not applicable when we aim to reconstruct the PSD (and not the original analog signal). This paper introduces a distributed wideband power spectrum sensing using multiple SUs to first estimate the power spectrum of signals received from sources in a collaborative manner. Each SU samples the received signal at sub-Nyquist rate and reconstructs the local PSD estimate based on the received digital samples. The local PSD estimate is then exchanged between SUs based on the consensus approach without fusion center. Once convergence on the PSD is reached, the detection on the existence of PUs is conducted. We found that for a PU signal power of 4 mW, noise power of 1 mW, and Rayleigh fading with the variance of -1 dB, the probability of detection can be at least 0.9 for the probability of a false alarm of 0.1 if the number of SUs is at least 40 or the compression rate is at least 0.4.
{"title":"Distributed Compressive Power Spectrum Sensing for Cognitive Radio","authors":"I. M. A. Wiryawan, D. D. Ariananda, S. Wibowo","doi":"10.1109/IAICT59002.2023.10205543","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205543","url":null,"abstract":"In cognitive radio (CR) networks, secondary users (SUs) might be required to gauge a wide frequency band to find frequency holes that they can use for signal transmission. When this spectrum sensing process is conducted digitally, a high sampling rate might be needed to satisfy the Nyquist rate. However, the existence of the frequency holes can be concluded by simply constructing the power spectral density (PSD) instead of the original signal. In fact, the Nyquist criterion is not applicable when we aim to reconstruct the PSD (and not the original analog signal). This paper introduces a distributed wideband power spectrum sensing using multiple SUs to first estimate the power spectrum of signals received from sources in a collaborative manner. Each SU samples the received signal at sub-Nyquist rate and reconstructs the local PSD estimate based on the received digital samples. The local PSD estimate is then exchanged between SUs based on the consensus approach without fusion center. Once convergence on the PSD is reached, the detection on the existence of PUs is conducted. We found that for a PU signal power of 4 mW, noise power of 1 mW, and Rayleigh fading with the variance of -1 dB, the probability of detection can be at least 0.9 for the probability of a false alarm of 0.1 if the number of SUs is at least 40 or the compression rate is at least 0.4.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"1 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":"130532033","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/IAICT59002.2023.10205374
Achmad Rizal Danisya, G. Hendrantoro, P. Handayani
In this paper, Grid Assisted Affinity Propagation Clustering (GAPC) algorithm is proposed to enhance the total spectral efficiency of Cell Head based Virtual Small Cell (CHVSC) service by increasing the number of Cell Heads (CH). The algorithm builds upon the previous method of Modified Affinity Propagation Clustering (MAPC) with addition of grid zonal division and Depth-First Search algorithm for advanced eligible-UE selection. Afterwards, both SNR and SIR are used for member selection in GAPC-SNR and GAPC-SIR respectively. From Monte Carlo simulation, MAPC still have higher average SINR compared to GAPC-SNR, but GAPC algorithm outperforms the MAPC algorithm in the number of CH appointed. With the compensation of higher accuracy of cluster finding inside MBS service zone, GAPC-SNR enhances overall bandwidth efficiency, silhouette score, and reduces computational complexity, as well as alleviating traffic burdens for each CH in comparison to MAPC.
本文提出了网格辅助亲和传播聚类(GAPC)算法，通过增加Cell Head (CH)的数量来提高基于Cell Head的虚拟小Cell (CHVSC)业务的总频谱效率。该算法在改进的关联传播聚类(MAPC)方法的基础上，增加了网格分区和深度优先搜索算法，用于高级的合格ue选择。然后在GAPC-SNR和GAPC-SIR中分别使用信噪比和SIR进行成员选择。从蒙特卡罗仿真来看，MAPC算法的平均信噪比仍然高于GAPC- snr，但GAPC算法在CH指定数量上优于MAPC算法。与MAPC相比，GAPC-SNR在补偿MBS服务区内更高的聚类查找精度的同时，提高了整体带宽效率和轮廓评分，降低了计算复杂度，减轻了每个CH的流量负担。
{"title":"UE Clustering Based on Grid Affinity Propagation for mmWave D2D in Virtual Small Cells","authors":"Achmad Rizal Danisya, G. Hendrantoro, P. Handayani","doi":"10.1109/IAICT59002.2023.10205374","DOIUrl":"https://doi.org/10.1109/IAICT59002.2023.10205374","url":null,"abstract":"In this paper, Grid Assisted Affinity Propagation Clustering (GAPC) algorithm is proposed to enhance the total spectral efficiency of Cell Head based Virtual Small Cell (CHVSC) service by increasing the number of Cell Heads (CH). The algorithm builds upon the previous method of Modified Affinity Propagation Clustering (MAPC) with addition of grid zonal division and Depth-First Search algorithm for advanced eligible-UE selection. Afterwards, both SNR and SIR are used for member selection in GAPC-SNR and GAPC-SIR respectively. From Monte Carlo simulation, MAPC still have higher average SINR compared to GAPC-SNR, but GAPC algorithm outperforms the MAPC algorithm in the number of CH appointed. With the compensation of higher accuracy of cluster finding inside MBS service zone, GAPC-SNR enhances overall bandwidth efficiency, silhouette score, and reduces computational complexity, as well as alleviating traffic burdens for each CH in comparison to MAPC.","PeriodicalId":339796,"journal":{"name":"2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT)","volume":"84 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":"113939657","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：