CLASSIFICATION OF HAND DRAWN GEOMETRIC SHAPES WITH MACHINE LEARNING METHODS

Creative Commons License

Kaya V., Arslan C.

3RD INTERNATIONAL AZERBAIJAN CONGRESS ON LIFE, ENGINEERING, AND APPLIED SCIENCES, Baku, Azerbaycan, 26 - 28 Kasım 2022, ss.111-116

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Basıldığı Şehir: Baku
  • Basıldığı Ülke: Azerbaycan
  • Sayfa Sayıları: ss.111-116
  • Erzincan Binali Yıldırım Üniversitesi Adresli: Evet

Özet

Today, computer vision and image processing techniques are used in many fields and these techniques are needed in different fields every day. Classification of hand-drawn geometric shapes is one of these areas. This study, it is aimed to correctly classify hand-drawn geometric shapes using machine learning methods. For this purpose, 7 different hand-drawn geometric shapes such as diagonal angle cross, ellipse, hexagon, line, square, straight cross, and triangle were studied. Many experiments were carried out by applying machine learning methods such as K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Logistic Regression to the images in this dataset. According to the results of these experiments, the SVM method showed more successful results than other machine learning methods. Considering the successful results obtained, it has been seen that it is possible to successfully classify hand-drawn geometric shapes using machine learning methods.