With the acceleration of globalization, cultural exchanges between China and the West have become more and more frequent, and English literature has been widely introduced into China. In this paper, the visual symbols in English literature are studied. In the pre-processing process of literary images, tilt correction, text region extraction and segmentation are carried out on the scanned images in turn. Based on DBNet algorithm, a text detection algorithm SMA for English literature is proposed, and Res2Net is used as a feature extraction network to extract multi-scale features to realize text extraction in literary images. Using the literary image preprocessing method and text extraction method proposed in this paper, we obtain the text of the novel “The Picture of Dorian Gray” by the British writer Oscar Wilde, and explore the visual symbols embedded in it. “Life”, “soul”, and “love” are the three keywords with the highest word frequency in the novel, which reach respectively 385, 366, and 340, implying the novel’s thematic information and visual symbols to a greater extent. By analyzing the visual symbols of the three main keywords of “life”, “soul”, and “love”, the novel “The Picture of Dorian Gray” is the crystallization of the fusion of Wilde’s moral and artistic views.
