Authors: Gang Liang
Addresses: Department of Art, School of Art and Design, Taizhou University, Taizhou 318000, China
Abstract: In order to overcome the problems of unreasonable contrast and low classification accuracy in traditional art image classification, this paper proposes a new art image classification method based on multiple naive Bayes algorithm. This method uses image smoothing method to remove the noise of the artwork images and realise image preprocessing. It constructs colour histogram to quantify the hue value of the artwork images to complete image feature extraction. The method obtains image feature points through multiple naive Bayes, then obtains the generated feature descriptor of the artwork images and finally, realises the classification of the artwork images. The experimental results show that the classification accuracy of this method is as high as 99.6%, the image classification contrast is always in the best value, and the classification time is short.
Keywords: multiple naive Bayes algorithm; artworks; image classification; image smoothing method; feature extraction; image feature descriptor.
International Journal of Arts and Technology, 2021 Vol.13 No.2, pp.79 - 90
Received: 23 Nov 2020
Accepted: 19 Apr 2021
Published online: 21 Jan 2022 *