Stroke Annotation of Ancient Characters Based on Crowdsourcing
Xiaofan Wang, Jing Xiong, Zhaoan Dong, Zhibin Liu, Wanli Huang, Boyong Wang · 2024
The decomposition and recombination of strokes play a crucial role in the evolution of ancient Chinese characters. In Oracle Bone Inscriptions (OBI), different semantic components carry distinct meanings, reflecting the structural patterns of character evolution and their cultural significance. Accurate extraction of strokes can effectively improve the accuracy of character recognition. The current Crowd-OBIGA crowdsourcing system only relies on key points annotation and cannot fully restore the complete stroke structure of OBI characters, which limits the effectiveness of glyph reconstruction. To solve this problem, this article extends the functionality of the crowdsourcing interaction system based on crowdsourcing technology, adding the function of grouping key points of OBI characters, in order to achieve more refined structured processing of character strokes. This improvement provides a rich data foundation for further decomposition and analysis of semantic components in OBI.