Browsing by Author "Gerdan, Dilara"

Now showing 1 - 3 of 3
  • No Thumbnail Available
    Item
    Görüntü işleme teknikleri kullanılarak bazı meyvelerin sınıflandırılması
    (Ankara : Ankara Üniversitesi : Fen Bilimleri Enstitüsü : Ziraat Fakültesi : Tarım Makineleri ve Teknolojileri Mühendisliği Ana Bilim Dalı, 2020) Gerdan, Dilara; Vatandaş, Mustafa; Ziraat Fakültesi
    Bu tez çalışmasında, meyvelerin boyut ve renk özelliklerine göre sınıflandırılması amacıyla bir görüntü işleme algoritması sınıflandırma düzeni geliştirilmiştir. Bu amaçla meyve çeşidi olarak, Starkrimson Delicious ve Golden Delicious elma çeşitleri, Washington Navel ve Valencia Midknight portakal çeşitleri ile Ekmek ve Eşme ayva çeşitlerinden 50'şer örnek seçilerek toplam 300 meyve denemeye alınmıştır. Kumpas ve spektrofotometre ile okunan boyut ve renk değerleri, geliştirilen görüntü işleme algoritmasına girilerek meyveleri doğru sınıflama başarısı belirlenmiştir. Görüntü işleme algoritmasının sınıflandırma ünitesi entegrasyonuyla, meyveleri boyutlarına göre sınıflandırma başarısı Starkrimson Delicious çeşidinde % 88, Golden Delicious elma çeşidinde % 100, Washington Navel çeşidinde % 96, Valencia Midknight çeşidinde % 82 bulunmuştur. Ayva sınıflandırma işleminde, TS 1817' de sınıflandırmanın ağırlıklara göre yapıldığı göz önüne alınarak her bir ayva hassas terazide tartılmış, boyutlar ile ağırlık arasındaki ilişkiler ortaya konulmuştur. Sınıflandırma işlemi, her iki çeşit için en küçük ve en büyük çap değerlerinin belirlenip, algoritmaya kumpasla okunan çap ölçülerinin alt ve üst limit değerleri girilerek yapılmıştır. Bu yöntemle, Ekmek ayva çeşidinde % 95, Eşme ayva çeşidinde ise sınıflama başarısı % 86 bulunmuştur. Her bir renk kanalı için, spektrofotometreden alınan alt ve üst limit değerleri algoritmaya girildiğinde elma çeşitlerinin birbirleri arasında renk bakımından sınıflara ayrılmasında başarı elma çeşitleri içinde % 100 bulunmuştur. Diğer taraftan görüntü işleme algoritmasıyla meyvelerde okunan boyut ve renk değerleri, veri madenciliğinde kullanılan tahminleyici teknikler kullanılarak değerlendirilmiştir. Bu amaçla, K En Yakın Komşuluk (KNN), Karar Ağacı (DT), Naive Bayes sınıflandırma ve Çok Katmanlı Algılayıcı Sinir Ağı (MLP) algoritmalarından yararlanılmıştır. Algoritmalar 10-kez çapraz doğrulama yöntemi ile çalıştırılmıştır. Ayrıca üst öğrenme algoritmalarından Rastgele Orman (RO) yöntemi seçilmiştir. Yapay sınıflandırıcıların eğitilerek denenmesinde ise boyut ve renk ölçümlerinin doğru meyve sınıfını tahminleme başarısı, KNN için % 93,6, DT için % 90,3, Naive Bayes % 88,3, MLP % 92,6 ve RO için % 94,3 bulunmuştur. In this thesis, an image processing algorithm and classification unit were developed to classify the fruits according to their size and color characteristics. For this purpose, a total of 300 fruits (50 fruit samples from each of the Starkrimson Delicious and Golden Delicious apple varieties, Washington Navel and Valencia Midknight orange varieties, Ekmek and Eşme quince varieties) were used in the experiments. The size and color values measured with a caliper and a spectrophotometer were entered in the developed image processing algorithm to determine the success rates of classifying the fruits. The integration of image processing algorithm with the classification unit classified 88% of the Starkrimson Delicious variety, 100% of the Golden Delicious apple variety, 96% of the Washington Navel variety, and 82% of the Valencia Midknight variety successfully. In the quince classification process, taking into consideration that the classification was made according to the weights in TS 1817 standard, each quince was weighed on a precision scale and the relationship between dimension and weight was determined. The smallest and largest diameters for both quince varieties were determined, then the highest and lowest diameters of each fruit were entered in the algorithm for classification. The success rates of classification with this method were found to be 95% for Ekmek and 86% for Eşme quince varieties, respectively. For each color channel, when the upper and lower limit values from the spectrophotometer were entered in the algorithm, the classification success was found to be 100% for the apple varieties. On the other hand, the size and color values of fruits with image processing algorithm were also evaluated by using estimation techniques in data mining. For this purpose, K Nearest Neighbor (KNN), Decision Tree (DT), Naive Bayes classification and Multi-Layer Perceptron Neural Network (MLP) algorithms were used. The algorithms were run using 10-fold cross-validation method. In addition, Random Forest (RO) method was chosen from the meta learning algorithms. The successes of predicting the correct fruit class and color measurements in training and testing of artificial classifiers were 93.6% for KNN, 90.3% for DT, 88.3% for Naive Bayes, 92.6% for MLP and 94.3% for RO, respectively.
  • No Thumbnail Available
    Item
    Görüntü işleme teknikleri kullanılarak bazı meyvelerin sınıflandırılması
    (Eğitim Bilimleri Enstitüsü, 2020) Gerdan, Dilara; Vatandaş, Mustafa; Other
    In this thesis, an image processing algorithm and classification unit were developed to classify the fruits according to their size and color characteristics. For this purpose, a total of 300 fruits (50 fruit samples from each of the Starkrimson Delicious and Golden Delicious apple varieties, Washington Navel and Valencia Midknight orange varieties, Ekmek and Eşme quince varieties) were used in the experiments. The size and color values measured with a caliper and a spectrophotometer were entered in the developed image processing algorithm to determine the success rates of classifying the fruits. The integration of image processing algorithm with the classification unit classified 88% of the Starkrimson Delicious variety, 100% of the Golden Delicious apple variety, 96% of the Washington Navel variety, and 82% of the Valencia Midknight variety successfully. In the quince classification process, taking into consideration that the classification was made according to the weights in TS 1817 standard, each quince was weighed on a precision scale and the relationship between dimension and weight was determined. The smallest and largest diameters for both quince varieties were determined, then the highest and lowest diameters of each fruit were entered in the algorithm for classification. The success rates of classification with this method were found to be 95% for Ekmek and 86% for Eşme quince varieties, respectively. For each color channel, when the upper and lower limit values from the spectrophotometer were entered in the algorithm, the classification success was found to be 100% for the apple varieties. On the other hand, the size and color values of fruits with image processing algorithm were also evaluated by using estimation techniques in data mining. For this purpose, K Nearest Neighbor (KNN), Decision Tree (DT), Naive Bayes classification and Multi-Layer Perceptron Neural Network (MLP) algorithms were used. The algorithms were run using 10-fold cross-validation method. In addition, Random Forest (RO) method was chosen from the meta learning algorithms. The successes of predicting the correct fruit class and color measurements in training and testing of artificial classifiers were 93.6% for KNN, 90.3% for DT, 88.3% for Naive Bayes, 92.6% for MLP and 94.3% for RO, respectively.
  • No Thumbnail Available
    Item
    Meta-Learning-Based Prediction of Different Corn Cultivars from Color Feature Extraction
    (Ankara Üniversitesi, 2021-03-04) Beyaz, Abdullah; Gerdan, Dilara; Ziraat Fakültesi
    Image analysis techniques are developing as applicable to the approaches of quantitative analysis, which is aimed to determine cultivar grains. Additionally, corn (Zea mays) grain processing companies evaluate the quality of kernels to determine the price of these cultivars. Because of this reason, in the study, a computer image analysis technique was applied on three corn cultivars. These were Zea mays L. indentata, Zea mays L. saccharata and a hybrid corn (Yellow sweet corn). These cultivars are commercially important as dry grains in Turkey. In the study, the grain color values were tested in the cultivars from Turkey’s collection. One hundred samples were used for each corn cultivar, and 300 corn grains in total were used for evaluations. Each of nine color parameters (Rmin, Rmean, Rmax, Gmin, Gmean, Gmax, Bmin, Bmean, Bmax) which were obtained from original RGB color channels with maximum and minimum values was evaluated from the digital images of three different corn cultivar grains. The values were analyzed with the help of the Multilayer Perceptron (MLP), Decision Tree (DT), Gradient Boost Decision Tree (GBDT) and Random Forest (RF) algorithms by using the Knime Analytics Platform. The majority voting method was applied to MLP and DT for prediction fusion. All algorithms were run with a 10-fold cross-validation method. The success of prediction accuracy was found as 99% for RF and GBDT, 97.66% for MLP, 96.66% DT and 97.40% for Majority Voting (MAVL). The MAVL method increased the accuracy of DT while decreasing the accuracy of MLP partly for the fusion of MLP and DT.