— Nowadays, people need easy access to basic nutrients to live a healthy life. In addition to providing calories that can meet the physiological needs of human beings, maize, which is one of the basic foods, contains valuable minerals and vitamins such as vitamin B6, sodium, magnesium, zinc, potassium, calcium, vitamin A. As a result of the increase in the world population in the world and our country, the need for maize is increasing day by day. Herein, it is important to detect the diseases seen in maize leaves that reduce the efficiency of maize production. Thanks to the developing technologies, producers should be encouraged by using technological opportunities in maize cultivation. In the study, it is aimed to detect maize rust, gray leaf spot, and leaf blight on maize leaves. In addition, two models based on the EfficientNetB5 network and convolutional neural network have been developed to detect diseases found in maize leaves using deep learning methods. To increase the performance metrics of created models, the number of images has been increased by using data augmentation techniques (mirror, rotation, scale). From the results, it is seen that the prediction success rates obtained in the EfficientNetB5 transfer learning model and the developed deep learning model are equal to 92.12% and 89.88%, respectively.
{"title":"Derin öğrenme ağları kullanılarak mısır yapraklarında hastalık tespiti","authors":"M. Göksu, Kubilay Muhammed Sünnetci, Ahmet Alkan","doi":"10.53070/bbd.989305","DOIUrl":"https://doi.org/10.53070/bbd.989305","url":null,"abstract":"— Nowadays, people need easy access to basic nutrients to live a healthy life. In addition to providing calories that can meet the physiological needs of human beings, maize, which is one of the basic foods, contains valuable minerals and vitamins such as vitamin B6, sodium, magnesium, zinc, potassium, calcium, vitamin A. As a result of the increase in the world population in the world and our country, the need for maize is increasing day by day. Herein, it is important to detect the diseases seen in maize leaves that reduce the efficiency of maize production. Thanks to the developing technologies, producers should be encouraged by using technological opportunities in maize cultivation. In the study, it is aimed to detect maize rust, gray leaf spot, and leaf blight on maize leaves. In addition, two models based on the EfficientNetB5 network and convolutional neural network have been developed to detect diseases found in maize leaves using deep learning methods. To increase the performance metrics of created models, the number of images has been increased by using data augmentation techniques (mirror, rotation, scale). From the results, it is seen that the prediction success rates obtained in the EfficientNetB5 transfer learning model and the developed deep learning model are equal to 92.12% and 89.88%, respectively.","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44694138","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}
{"title":"Hibrid 3B-2B ESA Mimarisi Kullanılarak Hiperspektral Uzaktan Algılama Görüntülerinin Sınıflandırılması","authors":"Hüseyin Fırat, M. Uçan, D. Hanbay","doi":"10.53070/bbd.989159","DOIUrl":"https://doi.org/10.53070/bbd.989159","url":null,"abstract":"","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44357661","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}
Hüseyin Üzen, İlhami Sel, Muammer Türkoğlu, D. Hanbay
Surface defect detection is one of the most important quality control components in manufacturing systems. The application of automatic surface defect detection methods in production systems is an important factor in ensuring high-quality products. In this study, depthwise separable convolution-based Deep Feature Pyramid Network (DÖPA) architecture was developed for automatic surface defect detection. In this network architecture, the learned parameters of the pre-trained VGG19 network architecture were used. MT dataset with defect detection images was used to test the performance of the proposed model. In experimental studies, 86.86% F1-score was obtained using the proposed DOPA architecture. These results showed that the proposed model was more successful than the existing studies.
{"title":"Derinlemesine Özellik Piramit Ağı Kullanarak Yüzey Hata Tespiti","authors":"Hüseyin Üzen, İlhami Sel, Muammer Türkoğlu, D. Hanbay","doi":"10.53070/bbd.990950","DOIUrl":"https://doi.org/10.53070/bbd.990950","url":null,"abstract":"Surface defect detection is one of the most important quality control components in manufacturing systems. The application of automatic surface defect detection methods in production systems is an important factor in ensuring high-quality products. In this study, depthwise separable convolution-based Deep Feature Pyramid Network (DÖPA) architecture was developed for automatic surface defect detection. In this network architecture, the learned parameters of the pre-trained VGG19 network architecture were used. MT dataset with defect detection images was used to test the performance of the proposed model. In experimental studies, 86.86% F1-score was obtained using the proposed DOPA architecture. These results showed that the proposed model was more successful than the existing studies.","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46103596","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}
In this study, the parameters of the distribution functions were adjusted with the optimization to optimization approach to improve the performance of the distribution function-based monarch butterfly optimization algorithm (MBO). For this, the random number generation processes, which greatly affect the flow of stochastic algorithms, were examined and the effect of distribution functions on these processes was determined. Then, the importance of parameter selection in the operation of distribution functions has been determined. It has been seen that the distribution function will be more effective with appropriate parameter selections. At this point, the distribution functions that can be used in the random number generation in the main target algorithm were tried to be determined with appropriate parameters with an upper auxiliary optimization algorithm. In conclusion; with the approach of optimization to optimization, the performance of the target algorithm has been tried to be increased and concrete results are presented in comparison with the tests made on the most used benchmark functions in the literature.
{"title":"Optimizasyonun Optimizasyonu Yaklaşımıyla Dağılım Fonksiyonu Tabanlı Kral Kelebeği Optimizasyon Algoritmasının Performansının Artırılması","authors":"Mehmet Akpamukçu, Abdullah Ateş","doi":"10.53070/bbd.990245","DOIUrl":"https://doi.org/10.53070/bbd.990245","url":null,"abstract":"In this study, the parameters of the distribution functions were adjusted with the optimization to optimization approach to improve the performance of the distribution function-based monarch butterfly optimization algorithm (MBO). For this, the random number generation processes, which greatly affect the flow of stochastic algorithms, were examined and the effect of distribution functions on these processes was determined. Then, the importance of parameter selection in the operation of distribution functions has been determined. It has been seen that the distribution function will be more effective with appropriate parameter selections. At this point, the distribution functions that can be used in the random number generation in the main target algorithm were tried to be determined with appropriate parameters with an upper auxiliary optimization algorithm. In conclusion; with the approach of optimization to optimization, the performance of the target algorithm has been tried to be increased and concrete results are presented in comparison with the tests made on the most used benchmark functions in the literature.","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43550653","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}
{"title":"Derin Öğrenme Yöntemleri ile Sıcaklık Tahmini: Diyarbakır İli Örneği","authors":"Aynur Sevinç, Buket Kaya","doi":"10.53070/bbd.990966","DOIUrl":"https://doi.org/10.53070/bbd.990966","url":null,"abstract":"","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42792299","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}
M. Güçlü, Burak Besceli, E. Polat, Timurhan Devellioglu, Gökberk Tamer, Nuh Mehmet Küçükusta, Enver Faruk Tanrikulu, A. Özen
{"title":"A Comparative Performance Evaluations of SC and MC VLC Systems in Underwater Environments","authors":"M. Güçlü, Burak Besceli, E. Polat, Timurhan Devellioglu, Gökberk Tamer, Nuh Mehmet Küçükusta, Enver Faruk Tanrikulu, A. Özen","doi":"10.53070/bbd.990734","DOIUrl":"https://doi.org/10.53070/bbd.990734","url":null,"abstract":"","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45535363","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}
— The demand for electrical energy in the world is increasing day by day. It is of great importance to make long-term electricity production forecasts in terms of meeting the increasing demand and determining the economic value of the generation investments that will be planned to be realized. In this study, an artificial neural network (ANN) based estimation methodology is presented for energy production forecasting using energy indicators such as Turkey's installed power capacity, gross electricity production
{"title":"Yapay Sinir Ağı (YSA) Kullanarak Farklı Kaynaklardan Türkiye’de Elektrik Enerjisi Üretim Potansiyelinin Tahmini","authors":"Harun Işık, Mustafa Şeker","doi":"10.53070/bbd.991039","DOIUrl":"https://doi.org/10.53070/bbd.991039","url":null,"abstract":"— The demand for electrical energy in the world is increasing day by day. It is of great importance to make long-term electricity production forecasts in terms of meeting the increasing demand and determining the economic value of the generation investments that will be planned to be realized. In this study, an artificial neural network (ANN) based estimation methodology is presented for energy production forecasting using energy indicators such as Turkey's installed power capacity, gross electricity production","PeriodicalId":41917,"journal":{"name":"Computer Science-AGH","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2021-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45819829","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：