Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786391
Qi Li, Jian Qu
Crowdfunding provides financial support to realize projects through the power of the public. Since crowdfunding is a voluntary activity, fraudulent crowdfunding projects cannot be regulated. During our research on how to detect fake crowdfunding projects, we found that the classification of product categories for crowdfunding project is crucial. Therefore, we want to implement the classification of product categories for crowdfunding projects on detecting fraudulent crowdfunding projects through the product ontology construction of crowdfunding projects.In this research, we proposed a novel method for product ontology construction based on the modified Nice Classification. To make the categories of products in the Nice Classification more closely match the crowdfunding projects, we have modified the Nice Classification according to the actual categories of products for crowdfunding projects. The method of product ontology construction based on the modified Nice Classification achieved an accuracy of 98% in classifying the categories of products for crowdfunding projects.
{"title":"Product Ontology Construction for Crowdfunding Projects","authors":"Qi Li, Jian Qu","doi":"10.1109/ICBIR54589.2022.9786391","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786391","url":null,"abstract":"Crowdfunding provides financial support to realize projects through the power of the public. Since crowdfunding is a voluntary activity, fraudulent crowdfunding projects cannot be regulated. During our research on how to detect fake crowdfunding projects, we found that the classification of product categories for crowdfunding project is crucial. Therefore, we want to implement the classification of product categories for crowdfunding projects on detecting fraudulent crowdfunding projects through the product ontology construction of crowdfunding projects.In this research, we proposed a novel method for product ontology construction based on the modified Nice Classification. To make the categories of products in the Nice Classification more closely match the crowdfunding projects, we have modified the Nice Classification according to the actual categories of products for crowdfunding projects. The method of product ontology construction based on the modified Nice Classification achieved an accuracy of 98% in classifying the categories of products for crowdfunding projects.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123835889","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786429
Farah Hana Kusumaputri, A. S. Arifin
The enormous internet development now day across all aspects of human life has introduced various hidden risk of malicious attacks on network security that most users didn’t realize. One of the malicious attacks is intrusion of system that proliferate user’s account effortlessly. Hence, in order to avoid intrusion effect that lead to financial loss and any other loss, intrusion detection system is needed to identify a dynamic pattern of cyber attacks. In this paper, we propose an Optimized XGBoost Classifier model with the help of Optuna Hypertuning method to find the best parameter for the model. In order to find the most efficient method for training, we assign three Optuna scenarios combine with feature selection to learn the data and the machine learning model. Through learning, Optuna generated the best parameter for XGBoost Classifier. Optuna avoids time consuming and low efficiency training model. The propose XGBoost Classifier model with Optuna Hypertuning method results in a greater accuracy of detection intrusion compare to any other models.
{"title":"Anomaly Detection based on NSL-KDD using XGBoost with Optuna Tuning","authors":"Farah Hana Kusumaputri, A. S. Arifin","doi":"10.1109/ICBIR54589.2022.9786429","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786429","url":null,"abstract":"The enormous internet development now day across all aspects of human life has introduced various hidden risk of malicious attacks on network security that most users didn’t realize. One of the malicious attacks is intrusion of system that proliferate user’s account effortlessly. Hence, in order to avoid intrusion effect that lead to financial loss and any other loss, intrusion detection system is needed to identify a dynamic pattern of cyber attacks. In this paper, we propose an Optimized XGBoost Classifier model with the help of Optuna Hypertuning method to find the best parameter for the model. In order to find the most efficient method for training, we assign three Optuna scenarios combine with feature selection to learn the data and the machine learning model. Through learning, Optuna generated the best parameter for XGBoost Classifier. Optuna avoids time consuming and low efficiency training model. The propose XGBoost Classifier model with Optuna Hypertuning method results in a greater accuracy of detection intrusion compare to any other models.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116304989","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786465
Voravarun Pattana-anake, Ferdin Joe John Joseph
Particulate Matter pollution with the magnitude of 2.5 microns is raising concerns from the most thickly populated cities around the world. There are various studies conducted on predictive analytics over the years. Deep learning has emerged as a new technology which is transforming the face of solving classification and regression problems. Various Long Short Term Memory based architectures are proposed in the past to predict time series data. Randomization of activation and optimization functions was done and the best performing combination is selected. Stack LSTM with this selected configuration on the PM2.5 data is found to be better than the existing LSTM based architectures. Inclusion of Adamax optimizer and fine tuning the activation functions in the LSTM layers gave better performance. The performance metrics reported in this paper are evident enough that the proposed architecture with optimized hyperparameters obtained by randomization of layers is found to perform with lesser error rates and training loss.
{"title":"Hyper Parameter Optimization of Stack LSTM Based Regression for PM 2.5 Data in Bangkok","authors":"Voravarun Pattana-anake, Ferdin Joe John Joseph","doi":"10.1109/ICBIR54589.2022.9786465","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786465","url":null,"abstract":"Particulate Matter pollution with the magnitude of 2.5 microns is raising concerns from the most thickly populated cities around the world. There are various studies conducted on predictive analytics over the years. Deep learning has emerged as a new technology which is transforming the face of solving classification and regression problems. Various Long Short Term Memory based architectures are proposed in the past to predict time series data. Randomization of activation and optimization functions was done and the best performing combination is selected. Stack LSTM with this selected configuration on the PM2.5 data is found to be better than the existing LSTM based architectures. Inclusion of Adamax optimizer and fine tuning the activation functions in the LSTM layers gave better performance. The performance metrics reported in this paper are evident enough that the proposed architecture with optimized hyperparameters obtained by randomization of layers is found to perform with lesser error rates and training loss.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121487751","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786495
Fatima Vapiwala, Deepika Pandita
The rapid technological shift caused by the pandemic in the field of education has compelled Indian educational institutions to adopt e-learning and e-assessment as a primary approach. The use of gamification software and technology the student assessment and evaluation plays a significant role in student engagement. A structured interview method was used for conducting this study and 200 responses were collected from post-graduate students through an interview process. The study provides significant insights into the crucial role of gamification not just in elearning but also in e-assessment of the students especially after the pandemic in the educational sector. The authors also propose a 5E model as a part of the strategy to be adopted by the Indian academicians and educators for utilizing gamification technology in e-learning and eassessment in the most beneficial way for the students.
{"title":"Strategies for Effective Use of Gamification Technology in E-Learning and E-Assessment","authors":"Fatima Vapiwala, Deepika Pandita","doi":"10.1109/ICBIR54589.2022.9786495","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786495","url":null,"abstract":"The rapid technological shift caused by the pandemic in the field of education has compelled Indian educational institutions to adopt e-learning and e-assessment as a primary approach. The use of gamification software and technology the student assessment and evaluation plays a significant role in student engagement. A structured interview method was used for conducting this study and 200 responses were collected from post-graduate students through an interview process. The study provides significant insights into the crucial role of gamification not just in elearning but also in e-assessment of the students especially after the pandemic in the educational sector. The authors also propose a 5E model as a part of the strategy to be adopted by the Indian academicians and educators for utilizing gamification technology in e-learning and eassessment in the most beneficial way for the students.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124064116","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786402
Vivat Thongchotchat, Kazuhiko Sato, H. Suto
Learning styles are preferences that individual learner has for learning information and responding to the learning environment. These styles can be used to help instructors design the courses to tailor each learner in order to make the learner learn effectively. The Index of Learning Styles (ILS) is the psychometric instrument for identifying Felder Silverman learning styles which have been used mainly, but there are many limitations of usages. This research aims to approach alternative ways to identify Felder-Silverman learning styles y using puzzle-game and machine learning. The acquired result shows that puzzle-game can be a practical alternative approach as learning styles identifier with weighted average accuracy 53.85% on procession styles, 84.615% on perception styles, and 53.85% on understanding styles compared to the result obtained from the ILS.
{"title":"Puzzle Game as Learning Identifier: HELLTAKER Use Case","authors":"Vivat Thongchotchat, Kazuhiko Sato, H. Suto","doi":"10.1109/ICBIR54589.2022.9786402","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786402","url":null,"abstract":"Learning styles are preferences that individual learner has for learning information and responding to the learning environment. These styles can be used to help instructors design the courses to tailor each learner in order to make the learner learn effectively. The Index of Learning Styles (ILS) is the psychometric instrument for identifying Felder Silverman learning styles which have been used mainly, but there are many limitations of usages. This research aims to approach alternative ways to identify Felder-Silverman learning styles y using puzzle-game and machine learning. The acquired result shows that puzzle-game can be a practical alternative approach as learning styles identifier with weighted average accuracy 53.85% on procession styles, 84.615% on perception styles, and 53.85% on understanding styles compared to the result obtained from the ILS.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126243923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786458
Jinjun Ruan, Jonathan M. Caballero, Ronaldo Juanatas
Combining the convolution neural network (CNN) model and bidirectional long short-term memory (BiLSTM) model, an ATT-CN-BILSTM Chinese news classification model is proposed based on the attention mechanism. The model uses the attention mechanism to improve the feature extraction process of CNN and BiLSTM. After cancelling the CNN pooling layer, it pays attention to the critical local features obtained by CNN convolution according to the timing features output by BiLSTM, giving full play to the respective advantages of CNN and BiLSTM models. The experimental results on Thucnews dataset show that the accuracy of the model for Chinese news text classification is 97.87%, and the recall rate and F1 score are better than the comparison model.
{"title":"Chinese News Text Classification Method Based On Attention Mechanism","authors":"Jinjun Ruan, Jonathan M. Caballero, Ronaldo Juanatas","doi":"10.1109/ICBIR54589.2022.9786458","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786458","url":null,"abstract":"Combining the convolution neural network (CNN) model and bidirectional long short-term memory (BiLSTM) model, an ATT-CN-BILSTM Chinese news classification model is proposed based on the attention mechanism. The model uses the attention mechanism to improve the feature extraction process of CNN and BiLSTM. After cancelling the CNN pooling layer, it pays attention to the critical local features obtained by CNN convolution according to the timing features output by BiLSTM, giving full play to the respective advantages of CNN and BiLSTM models. The experimental results on Thucnews dataset show that the accuracy of the model for Chinese news text classification is 97.87%, and the recall rate and F1 score are better than the comparison model.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126327076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786413
Sawittree Jumpathong, T. Theeramunkong, T. Supnithi, M. Okumura
This paper presents a performance analysis of deep-learning-based Thai news abstractive summarization. The analysis focuses on the position of the words in the original document that are generated into the summary. Also, the analysis includes the behavior of word generation of the system. Moreover, we analyse how the document length affects the performance of the models regarding word positions of the original document. The result of the experiment shows that the models generated the output summary by generating most words from the beginning part more than those from the reference summary about 1.79 times on the TR testing dataset and about 2.03 times on the TPBS testing dataset. Additionally, the models occasionally generated words that do not exist in the original document about 1.68% of word number of the summary on the TR testing dataset and about 0.88% of word number of the summary on the TBPS testing dataset. According to the result, it is found that the models generated words in the system summary is not consistent with words in the reference summary. In the document length, it is found that the models can summarize a short document better than a long document.
{"title":"A Performance Analysis of Deep-Learning-Based Thai News Abstractive Summarization: Word Positions and Document Length","authors":"Sawittree Jumpathong, T. Theeramunkong, T. Supnithi, M. Okumura","doi":"10.1109/ICBIR54589.2022.9786413","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786413","url":null,"abstract":"This paper presents a performance analysis of deep-learning-based Thai news abstractive summarization. The analysis focuses on the position of the words in the original document that are generated into the summary. Also, the analysis includes the behavior of word generation of the system. Moreover, we analyse how the document length affects the performance of the models regarding word positions of the original document. The result of the experiment shows that the models generated the output summary by generating most words from the beginning part more than those from the reference summary about 1.79 times on the TR testing dataset and about 2.03 times on the TPBS testing dataset. Additionally, the models occasionally generated words that do not exist in the original document about 1.68% of word number of the summary on the TR testing dataset and about 0.88% of word number of the summary on the TBPS testing dataset. According to the result, it is found that the models generated words in the system summary is not consistent with words in the reference summary. In the document length, it is found that the models can summarize a short document better than a long document.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115837908","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786392
Denuxshe Chelvaratnam, Aysha Nazar, Sangeetha Balasingam, T. Kartheeswaran
One of the most vital aspects of our lives is transportation. Mainly, everyone relies upon road transportation to meet their fundamental daily needs and respond to emergencies. The minimum requirement for a good vehicle is the road surface, also known as road quality. However, most roads presently have rough and uneven surfaces, such as pits and bumps. This is a crucial problem for transportation in some underdeveloped countries. We discussed a potentially simple, inexpensive methodology to fix this problem utilizing the sensors of the very commonly used smartphone by everyone in the community. We can do this with the help of GPS and the accelerometer of smartphones. With the help of the shake threshold of the smartphone accelerometer, we can determine the road condition and acquire longitude and latitude details from the GPS sensor to find the location. The readings obtained from the smartphone can be collected through the dedicated app and used to measure the quality of the selected road. The selected three paths will be re-ranked based on these measurements.
{"title":"Quality Based Road Ranking System Using Accelerometer and GPS of Sensors of Smartphones","authors":"Denuxshe Chelvaratnam, Aysha Nazar, Sangeetha Balasingam, T. Kartheeswaran","doi":"10.1109/ICBIR54589.2022.9786392","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786392","url":null,"abstract":"One of the most vital aspects of our lives is transportation. Mainly, everyone relies upon road transportation to meet their fundamental daily needs and respond to emergencies. The minimum requirement for a good vehicle is the road surface, also known as road quality. However, most roads presently have rough and uneven surfaces, such as pits and bumps. This is a crucial problem for transportation in some underdeveloped countries. We discussed a potentially simple, inexpensive methodology to fix this problem utilizing the sensors of the very commonly used smartphone by everyone in the community. We can do this with the help of GPS and the accelerometer of smartphones. With the help of the shake threshold of the smartphone accelerometer, we can determine the road condition and acquire longitude and latitude details from the GPS sensor to find the location. The readings obtained from the smartphone can be collected through the dedicated app and used to measure the quality of the selected road. The selected three paths will be re-ranked based on these measurements.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131377445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786515
Wongsathon Angkhem, S. Tantrairatn
The Unmanned Aerial Vehicle is an effective vehicle for rescue, search, and surveillance missions. A thermal camera improves the UAV system to operate these missions in the nighttime. Real-time human detection is an algorithm to increase performance and improve to be fully autonomous in rescue missions. Many studies have led to the integration of realtime human detection from thermal aerial images, but the task remains difficult from various human features from multi capture angle and UAV altitude. This paper proposes an experimental process for implementing real-time human detection from UAVs in the nighttime. We choose the YOLOv3 model for real-time human detection. Then, we create a custom thermal aerial human dataset that multi-capturing angle and altitude. The dataset is captured in the same condition of UAVs operation. We prepare and preprocess the dataset before sending it to the model training process. Finally, we evaluate a trained model for mean-Average Precision. The accuracy of prediction is evaluated with a test set and real-time detection performance. The results demonstrate that the model can detect a human in real-time with a thermal image from a UAV view and the accuracy of detection is mAP of 64.8% in the operating range of the UAV.
{"title":"Night-Time Human Detection From UAV","authors":"Wongsathon Angkhem, S. Tantrairatn","doi":"10.1109/ICBIR54589.2022.9786515","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786515","url":null,"abstract":"The Unmanned Aerial Vehicle is an effective vehicle for rescue, search, and surveillance missions. A thermal camera improves the UAV system to operate these missions in the nighttime. Real-time human detection is an algorithm to increase performance and improve to be fully autonomous in rescue missions. Many studies have led to the integration of realtime human detection from thermal aerial images, but the task remains difficult from various human features from multi capture angle and UAV altitude. This paper proposes an experimental process for implementing real-time human detection from UAVs in the nighttime. We choose the YOLOv3 model for real-time human detection. Then, we create a custom thermal aerial human dataset that multi-capturing angle and altitude. The dataset is captured in the same condition of UAVs operation. We prepare and preprocess the dataset before sending it to the model training process. Finally, we evaluate a trained model for mean-Average Precision. The accuracy of prediction is evaluated with a test set and real-time detection performance. The results demonstrate that the model can detect a human in real-time with a thermal image from a UAV view and the accuracy of detection is mAP of 64.8% in the operating range of the UAV.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131625081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-19DOI: 10.1109/ICBIR54589.2022.9786519
Yanusha Mehendran, T. Kartheeswaran, N. Kodikara
Bananas provide rapid energy and are a worldwide available fruit. Bananas are also available all year and seldom cause health problems. Banana is one of the most significant fruits in Sri Lanka since it is extensively consumed and suitable for all situations. Bananas are undoubtedly healthful and have export worth. As a result, determining freshness is critical to ensuring product quality and market value. The conventional method for measuring the freshness of a banana in terms of days necessitates the naked eye inspection of experienced specialists. Because specialists are not always available, we developed a method for determining the freshness of bananas using image processing techniques. For this investigation, images of bananas at various levels were obtained using a high-quality mobile camera. K-Means clustering was used to identify the interesting region of the bananas, and a Support Vector Machine (SVM) model was utilized to estimate freshness by training using the chosen features from the input images. Several feature combinations were investigated for this study’s evaluation, and the relationship between the features Energy, Contrast, Correlation, RMS, Homogeneity, Mean, Standard deviation, Entropy, Greenness, Kurtosis, Skewness, and Variance yielded an accuracy of 81.75 percent. This system’s goal is to inspire future researchers to turn this method into a mobile application that also incorporates Artificial Intelligence (AI) for autonomous bulk observation.
{"title":"Banana Freshness Identification Using Image Processing Techniques","authors":"Yanusha Mehendran, T. Kartheeswaran, N. Kodikara","doi":"10.1109/ICBIR54589.2022.9786519","DOIUrl":"https://doi.org/10.1109/ICBIR54589.2022.9786519","url":null,"abstract":"Bananas provide rapid energy and are a worldwide available fruit. Bananas are also available all year and seldom cause health problems. Banana is one of the most significant fruits in Sri Lanka since it is extensively consumed and suitable for all situations. Bananas are undoubtedly healthful and have export worth. As a result, determining freshness is critical to ensuring product quality and market value. The conventional method for measuring the freshness of a banana in terms of days necessitates the naked eye inspection of experienced specialists. Because specialists are not always available, we developed a method for determining the freshness of bananas using image processing techniques. For this investigation, images of bananas at various levels were obtained using a high-quality mobile camera. K-Means clustering was used to identify the interesting region of the bananas, and a Support Vector Machine (SVM) model was utilized to estimate freshness by training using the chosen features from the input images. Several feature combinations were investigated for this study’s evaluation, and the relationship between the features Energy, Contrast, Correlation, RMS, Homogeneity, Mean, Standard deviation, Entropy, Greenness, Kurtosis, Skewness, and Variance yielded an accuracy of 81.75 percent. This system’s goal is to inspire future researchers to turn this method into a mobile application that also incorporates Artificial Intelligence (AI) for autonomous bulk observation.","PeriodicalId":216904,"journal":{"name":"2022 7th International Conference on Business and Industrial Research (ICBIR)","volume":"906 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130744941","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：