A study on the influence of digital painting image quality on visual art and style classification Based on the sparse encoding algorithm

Abstract

At present, since the analysis of low-resolution paintings is less effective in studying the quality of painting images, the study of the impact of digital painting image quality on visual art and style classification is proposed. The high and low resolution images of four paintings are selected as the research objects, and the basis functions of the paintings are trained using sparse coding, and the features of the paintings are extracted using information theory to extract seven features such as Gabor energy, peak direction and peak space of the basis functions. Finally, the features extracted from the high-resolution painting images were used to classify the painting styles. The experimental results show that the features extracted from the low-resolution painting images still have the ability to characterize the painting style to a certain extent and can be used for the analysis of painting style.