Kernel Matrix Based Textile Image Categorization Using Machine Learning
Abstract views: 22
/ 
PDF downloads: 24
Keywords:
Textile Images, Support Vector Machines, Multi-Layer Perceptron, K-Nearest Neighbor, Kernel MatrixAbstract
Today, just as text and audio data are rapidly increasing, visual data is also growing at a fast
pace. In textile technology, fabric patterns, and images have encountered a similar expansion.
Consequently, resolving this large volume of data becomes a significant problem. We need efficient
classification to quickly access this data, and this categorization process should be automated using
machine learning techniques. In our study, we have constructed the systems for classifying textile images
using machine learning methods such as Multilayer Perceptron, Support Vector Machines, and K-Nearest
Neighbors. The textile dataset consists of color and grayscale images, divided into training and test sets.
Models are trained using the training data, and their decision-making performance is evaluated using the
test data. During model generation, preprocessing is performed first. All images are converted to black
and white. Edge detection filters like Sobel and Prewitt are applied to find the edges in the images.
Optionally, thinning can also be applied before this step. After preprocessing, feature extraction is carried
out. For each image, the frequency of matrices called kernel matrices, which slide over the image, is
calculated and normalized. This representation allows images to be transformed into vectors, which are
then used to train machine learning models. In the testing phase, commonly used metrics such as F-score
and Accuracy are employed to evaluate the performance of these systems. The developed models are
compared to each other, and the most successful methods are determined.
Downloads
References
R. Aşlıyan, “Classification of Textile Images”, Graduate School of Natural and Applied Sciences, Computer Engineering, Dokuz Eylül University, M.Sc. Thesis, İzmir, 2002.
R. Aşlıyan, and A. Alpkoçak, Tekstil Desenlerinin Otomatik Olarak Sınıflandırılması Üzerine Bir Çalışma. SİU2002. 10. Sinyal İşleme ve İletişim Uygulamaları Kurultayı. Cilt I s. 123-128, Pamukkale, Denizli, 2002.
R. Aşlıyan, “Textile Image Classification: Categorizing huge amout of textile images efficiently”. Lambert Academic Publishing AG& CO. KG. Saarbrücken, Germany. ISBN: 978-3-8383-5732-4. 2010.
İ. Ulvi, R. Aşlıyan and K. Günel “Textile Image Classification Using Artificial Neural Networks”, 3rd World Conference on Innovation and Computer Science ( INSODE - 2013 ), Antalya, Turkey, 2013.
T. Geyers, F. Aldershoff and A. W. M. Smeulders, “Classification of Images on Internet by Visual and Textual Information”, In SPIE Vol: 3964: doi: 10.1117/12.373453, San Jose, 2000.
C. J. C. Burges, “A Tutorial on Support Vector Machines for Pattern Recognition”, Knowledge Discovery and Data Mining: 2(2) , 1998.
O. G. Sezer, A. Erçil, M. Keskinöz, “Destek Vektör Makinesi Kullanarak Bağımsız Bileşen Tabanlı 3B Nesne Tanıma”, SUI 2005, Sabancı Üniversitesi Mühendislik ve Doğa Bilimleri Fakültesi, 2005.
A . Akkuş and H. A. Güvenir, “K-Nearest Neighbor Classification on Feature Projections”. In Proc. ICMI'96, Lorenzo Saitta (Ed.), Morgan Kaufmann, Bari, Italy, pp. 12-19, 1996.
Y. Özkan, Veri Madenciliği Yöntemleri, Dr. Rifat Çölkesen Dr. Cengiz Uğurkaya Papatya Yayıncılık, İstanbul, 2013.
S. Haykin, Neural Networks: A Comprehensive Foundation, Macmillan College Publishing Company, New York, 1994.
M. Ö. Efe and O. Kaynak, Artificial Neural Networks, Boğaziçi University Publishing, İstanbul, 2000.
E. Öztemel, Yapay Sinir Ağları, Papatya Yayıncılık Eğitim, Türkiye, 2012.
R. C. Gonzalez, R. Woods, Introduction. Digital Image Processing, 3rd Edition, Prentice-Hall, New Jersey, 2008.
E. Timur and C. Sarı, “Agora (Magnesia/Aydın) Manyetik Verilerinin Kenar Belirleme İşleçleri ve 3-Boyutlu Ters Çözümle Modellenmesi”, Hacettepe Üniversitesi Yerbilimleri Uygulama ve Araştırma Merkezi Dergisi, vol. 31(2), pp. 67-82, 2010.
I. Sobel and G. Feldman, A 3x3 Isotropic Gradient Operator For Image Processing, John Wiley and Sons, New York, US, 1968.
R. Boyle and R. Thomas, Computer Vision: A First Course, Blackwell Scientific Publications, pp. 48-50, 1988.
J. M. S. Prewitt, Object Enhancement and Extraction, Picture Processing and Psychopictorics, Editors Lipkin, B., Rosenfeld, A., Academic Press, New York, 1970.
S. Konishi, A. L. Yuille, J. M. Coughlan, S. C. Zhu, “Statistical Edge Detection: Learning and Evaluating Edge Cues”, IEEE Transactions on, Pattern Analysis and Machine Intelligence, vol. 25(1), pp. 57-74, 2003.
S. M. Özkul, “Tek Kamera ile Görüntüde Derinliğin Hesaplanması”, Osmangazi Üniversitesi Fen Bilimleri Enstitüsü Elektrik Elektronik Mühendisliği, Yüksek Lisans Tezi, Eskişehir, 1995.
S. Bilgi, “Çok Ölçekli Kartografik Gösterimlerde Mekansal Bilginin Nicelik Analizi”, İstanbul Teknik Üniversitesi Fen Bilimleri Enstitüsü Geomatik Mühendisliği Programı Gemomatik Mühendisliği Anabilimdalı, Doktora Tezi, sf. pp. 13-15, İstanbul, 2012.
J. J. Clark, “Authenticating Edges Produced by Zero-Crossing Algorithms”, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol 2(1), pp. 43-57, 1989
H. Blum, “A Transformation for Extracting New Descriptors of Shape, Models for the Perception of Speech and Visual Form”, MIT Press, Cambridge, pp. 362–380, 1967.
N. Ateş, “Destek Vektör Makineleri ve Gauss Karışım Modeli ile İstenmeyen E-postaların Tespiti”, Süleyman Demirel Üniversitesi Fen Bilimleri Enstitüsü Bilgisayar Mühendisliği Anabilim Dalı, Yüksek Lisans Tezi, Isparta, 2014.
S. Ayhan and Ş. Erdoğmuş, “Destek Vektör Makineleriyle Sınıflandırma Probleminin Çözümü İçin Çekirdek Fonksiyonu Seçimi”, Eskişehir Osmangazi Üniversitesi İİBF Dergisi, vol. 9(1), pp.175-198, 2014.
