Image Processing and Traditional Machine Learning Based Classification of Brown Marmorated Stink Bug (Halyomorpha Halys) Defected Hazelnut

Omsalma Alsadig Adam Gadalla; Y. Benal Öztekin

Makale detayları

Image Processing and Traditional Machine Learning Based Classification of Brown Marmorated Stink Bug (Halyomorpha Halys) Defected Hazelnut

Yazarlar:

Omsalma Alsadig Adam Gadalla | Y. Benal Öztekin

Yayıncı:

Tekirdag Namik Kemal University

Ana Başlıklar:

Tarım Makine Sistemleri

Dergi:

Tekirdağ Ziraat Fakültesi Dergisi Vol.20 Issue 4, pp.784-798

Belge Tipi:

Makale

Yayın Tarihi:

12/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001305433

pISSN:

1302-7050

Anahtar kelimeler:

Özet:

 Quality control of hazelnuts is a major concern in many regions across the world, but particularly in Turkey as the worlds largest hazelnut producer. Using image processing and deep learning techniques, this study intended to detect and classify healthy hazelnuts and hazelnuts infected with the Brown Marmorated Stink Bug. Infected hazelnut samples were collected from the 2021 production period by experts. A Guppy Pro CCD camera-based image acquisition system was used to capture hazelnut images. A total of 400 RGB hazelnut images were captured to train machine learning models. Image segmentation process was carried out to subtract hazelnut images from the background using the Thresholding technique. Moment features were extracted from RGB and lab spaces to be used to train traditional machine learning models. Furthermore, the most relevant and discriminative feature set was selected using the Boruta feature selection method. Traditional machine learning models including Random Forest, Support Vector Machine, Logistic Regression, Naive Bayes, and Decision Tree were trained twice, once with all features and another with the selected feature set only. The overall accuracy, statistical characteristics of the confusion matrix, and model training time were all calculated to evaluate and compare models performances. As a result, threshold value of 50 was determined from the gray level histogram and was able to separate hazelnut image from the background perfectly. Only seven moment features were identified as the most discriminative features out of 24 features. The SVM model with all feature vectors had the greatest classification accuracy of 98.75 . When only the selected features were employed, the performance of Random Forest and Logistic Regression models improved to 97.5 and 96.25 , respectively.

 Quality control of hazelnuts is a major concern in many regions across the world, but particularly in Turkey as the worlds largest hazelnut producer. Using image processing and deep learning techniques, this study intended to detect and classify healthy hazelnuts and hazelnuts infected with the Brown Marmorated Stink Bug. Infected hazelnut samples were collected from the 2021 production period by experts. A Guppy Pro CCD camera-based image acquisition system was used to capture hazelnut images. A total of 400 RGB hazelnut images were captured to train machine learning models. Image segmentation process was carried out to subtract hazelnut images from the background using the Thresholding technique. Moment features were extracted from RGB and lab spaces to be used to train traditional machine learning models. Furthermore, the most relevant and discriminative feature set was selected using the Boruta feature selection method. Traditional machine learning models including Random Forest, Support Vector Machine, Logistic Regression, Naive Bayes, and Decision Tree were trained twice, once with all features and another with the selected feature set only. The overall accuracy, statistical characteristics of the confusion matrix, and model training time were all calculated to evaluate and compare models performances. As a result, threshold value of 50 was determined from the gray level histogram and was able to separate hazelnut image from the background perfectly. Only seven moment features were identified as the most discriminative features out of 24 features. The SVM model with all feature vectors had the greatest classification accuracy of 98.75 . When only the selected features were employed, the performance of Random Forest and Logistic Regression models improved to 97.5 and 96.25 , respectively.

 Quality control of hazelnuts is a major concern in many regions across the world, but particularly in Turkey as the worlds largest hazelnut producer. Using image processing and deep learning techniques, this study intended to detect and classify healthy hazelnuts and hazelnuts infected with the Brown Marmorated Stink Bug. Infected hazelnut samples were collected from the 2021 production period by experts. A Guppy Pro CCD camera-based image acquisition system was used to capture hazelnut images. A total of 400 RGB hazelnut images were captured to train machine learning models. Image segmentation process was carried out to subtract hazelnut images from the background using the Thresholding technique. Moment features were extracted from RGB and lab spaces to be used to train traditional machine learning models. Furthermore, the most relevant and discriminative feature set was selected using the Boruta feature selection method. Traditional machine learning models including Random Forest, Support Vector Machine, Logistic Regression, Naive Bayes, and Decision Tree were trained twice, once with all features and another with the selected feature set only. The overall accuracy, statistical characteristics of the confusion matrix, and model training time were all calculated to evaluate and compare models performances. As a result, threshold value of 50 was determined from the gray level histogram and was able to separate hazelnut image from the background perfectly. Only seven moment features were identified as the most discriminative features out of 24 features. The SVM model with all feature vectors had the greatest classification accuracy of 98.75 . When only the selected features were employed, the performance of Random Forest and Logistic Regression models improved to 97.5 and 96.25 , respectively.

 Fındığın kalite kontrolü, dünyanın birçok bölgesinde, özellikle de dünyanın en büyük fındık üreticisi olan Türkiyede büyük bir problem kaynağıdır. Bu çalışma, görüntü işleme ve derin öğrenme tekniklerini kullanarak, Kahverengi Kokarca ile enfekte olmuş ve sağlıklı fındıkları birbirinden ayırarak belirlemek ve sınıflandırmak amaçlanmıştır. Kahverengi Kokarcalı fındık örnekleri, uzmanlar tarafından 2021 üretim döneminden elde edilmiştir. Fındık görüntülerini yakalamak için Guppy Pro CCD kamera tabanlı görüntü alma sistemi kullanılmıştır. Geleneksel makine öğrenme modellerini eğitmek için toplam olarak 400 RGB fındık görüntüsü alınmıştır. Fındık görüntülerinin arka plandan çıkarılması için görüntü bölüntüleme işlemi Eşikleme tekniği kullanılarak gerçekleştirilmiştir. Fındık moment özellikleri, geleneksel makine öğrenme modellerini eğitmek için kullanılmak üzere RGB ve lab renk çıkarılmıştır. Ayrıca, Boruta özellik seçim yöntemi kullanılarak en önemli ve en ayırt edici öznitelik seti seçilmiştir. Rastgele Orman, Destek Vektör Makinesi, Lojistik Regresyon, Naive Bayes ve Karar Ağacı dâhil olmak üzere geleneksel makine öğrenme modelleri, bir kez tüm özelliklerle ve bir kez daha yalnızca seçilmiş özelliklerle olmak üzere iki kez eğitilmiştir. Genel doğruluk, karışıklık matrisinin istatistiksel özellikleri ve model eğitim süresinin tümü, modelin sınıflandırma performansını değerlendirmek ve karşılaştırmak için hesaplanmıştır. Sonuç olarak, gri seviye histogramından 50 eşik değeri belirlenmiştir ve fındık görüntüsünü arka plandan mükemmel bir şekilde ayırabilmiştir. Çıkartılmış 24 özellik arasından en ayırt edici özellik olarak sadece yedi tane renk özelliği belirlenmiştir. Tüm çıkartılmış özellikler kullandıktan sonra Destek Vektör Makinesi modeli kullanılarak, 98.75 ile en yüksek sınıflandırma doğruluğu elde edilmiştir. Aynı zamanda tüm özelliklerden sadece seçilen özellikler kullanıldığında Rastgele Orman ve Lojistik Regresyon sınılandırıcılarının modellerinin performansı sırasıyla 97.5 ve 96.25e kadar yükselmiştir.

Makale detayları

Image Processing and Traditional Machine Learning Based Classification of Brown Marmorated Stink Bug (Halyomorpha Halys) Defected Hazelnut

Yazarlar:

Omsalma Alsadig Adam Gadalla | Y. Benal Öztekin

Yayıncı:

Tekirdag Namik Kemal University

Ana Başlıklar:

Tarım Makine Sistemleri

Dergi:

Tekirdağ Ziraat Fakültesi Dergisi Vol.20 Issue 4, pp.784-798

Belge Tipi:

Makale

Yayın Tarihi:

12/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001305433

pISSN:

1302-7050

Support Vector Machine

Hazelnut

Feature selection

Feature extraction

BorutaBoruta

Destek Vektör Makinesi

Fındık

Özellik seçme

Özellik çıkartma

Sayı

Makale detayları

Yazarlar: Omsalma Alsadig Adam Gadalla | Y. Benal Öztekin

Yayıncı: Tekirdag Namik Kemal University

Ana Başlıklar: Tarım Makine Sistemleri

Dergi: Tekirdağ Ziraat Fakültesi Dergisi Vol.20 Issue 4, pp.784-798

Belge Tipi: Makale

Yayın Tarihi: 12/2023

Belge Dili: İngilizce

Document ID (Org No.): 2001305433

pISSN: 1302-7050

Sayı

Yazarlar:

Omsalma Alsadig Adam Gadalla | Y. Benal Öztekin

Yayıncı:

Tekirdag Namik Kemal University

Ana Başlıklar:

Tarım Makine Sistemleri

Dergi:

Tekirdağ Ziraat Fakültesi Dergisi Vol.20 Issue 4, pp.784-798

Belge Tipi:

Makale

Yayın Tarihi:

12/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001305433

pISSN:

1302-7050