Calligraphy Character Detection Based on Deep Convolutional Neural Network

Calligraphy (the special art of drawing characters with a brush specially made by the Chinese) is an integral part Chinese culture, and detecting calligraphy highly significant. At present, there are still some challenges in detection ancient calligraphy. In this paper, we interested character problem focusing on boundary. We chose High-Resolution Net (HRNet) as feature extraction backbone network to learn reliable high-resolution representations. Then, used scale prediction branch spatial information detect region categorize its boundaries. channel attention mechanism fusion method improve effectiveness process. Finally, pre-trained self-generated database fine-tuned real database. set up two groups ablation studies for comparison, comparison results proved superiority our method. This paper found that classification boundaries has certain auxiliary effect single detection.