Separation of original paintings of matisse and his fakes using wavelet and artificial neural networks

Abstract

In recent years, with the latest developments in computer technology, wavelet and Artificial Neural Networks (ANN) are being widely used in different fields and disciplines. Especially, wavelet followed by ANN applications has produced successful results in image processing. In this paper, we have applied wavelet followed by ANN to obtain an objective approach in seperating original paintings of Matisse and fakes. In art environment, the term of fake can be explained as producing new paintings resembling to the artist's painting style. The works of Matisse have been especially chosen since his paintings are mostly faked. Here, wavelet is utilized for feature extraction of 2D paintings. Thus, important properties of input image are extracted while reducing input parameters with minimum loss of information. ANN is then applied in separation process between paintings. At the end of the overall separation task, we obtained 88 % classification accuracy.