Based on semiotics, this paper systematically researches the preprocessing, structural analysis and artistic information extraction methods of calligraphic characters for the key problems in the digitization of calligraphic art. Combining the writing rules and the definition of skeleton diagram to realize the text structure analysis, and put forward the algorithm of calligraphic art information extraction based on partitioned bootstrap filtering. Carrying out experiments on digitization of calligraphic art and introducing different mainstream methods to compare and verify their effectiveness. In the task of burr detection, this paper’s algorithm has the best overall performance, with an accuracy of 99.27%, which is significantly higher than that of the threshold method (97.32%), the significance-ordered pruning method (80.13%) and the automatic pruning method (94.92%). In the skeleton extraction task, this paper’s algorithm strokes the skeleton with an accuracy of 99.04% and a processing speed of 8.93 characters/second, striking a balance between accuracy and efficiency. In the quantitative comparison experiment this paper algorithm still has obvious improvement on the index, F1, IoU is 0.7018, 0.5264 respectively.
