增强难检字识别：手动扩充数据集并引入甲骨文识别辅助算法

Translation: We manually expanded the dataset to increase the overall data size and specifically added training data for difficult-to-detect characters, strengthening the detection model's focus on difficult-to-detect characters. We also conducted secondary labeling of the dataset to improve its credibility. To address the issue of misclassification of characters similar to oracle bone script, we used an oracle bone script recognition-assisted detection algorithm to filter out incorrect prediction results.