Computer and Information Technology Engineering Department, Faculty of Engineering, Razi University, Kermanshah, Iran
Abstract
Stylometry is one of the key issues in art work recognation, however most artists do not identify their styles. Generally, people often empirically recognize an artist's style through following the artist's paintings and paying attention to the paintings' details. This paper, for the first time, proposes an approach to classify Iranian painters' style utilising image processing techniques. For feature extraction, histogram of gradient (HOG) and local binary patterns (LBP) are exploited applying support vector machine (SVM) for classification. To assess the proposed method, one dataset of paintings that contains five famous Iranian painters, namely Hossein Behzad, Kamal-ol-Molk, Morteza Katouzian, Sohrab Sepehri and Mahmoud Farshchian, including 326 paintings, is collected. The experimental results indicate that our proposed method can well classify the painting styles, where, different styles are classified with average accuracy rate of 95.48%.
Keshvari, S., & Chalechale, A. (2017). Stylometry of Painting Using Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP). TABRIZ JOURNAL OF ELECTRICAL ENGINEERING, 47(3), 1195-1204.
MLA
S. Keshvari; A. Chalechale. "Stylometry of Painting Using Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP)". TABRIZ JOURNAL OF ELECTRICAL ENGINEERING, 47, 3, 2017, 1195-1204.
HARVARD
Keshvari, S., Chalechale, A. (2017). 'Stylometry of Painting Using Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP)', TABRIZ JOURNAL OF ELECTRICAL ENGINEERING, 47(3), pp. 1195-1204.
VANCOUVER
Keshvari, S., Chalechale, A. Stylometry of Painting Using Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP). TABRIZ JOURNAL OF ELECTRICAL ENGINEERING, 2017; 47(3): 1195-1204.
)