Context-Aware Chart Element Detection
This work addresses chart element detection, a domain-specific prerequisite for data extraction, with incremental improvements in context integration and categorization.
The paper tackles the problem of accurately detecting chart basic elements for chart data extraction by proposing CACHED, a context-aware method that integrates local-global context fusion with Cascade R-CNN, achieving state-of-the-art performance and the best result on the PMC test dataset for bar plot detection.
As a prerequisite of chart data extraction, the accurate detection of chart basic elements is essential and mandatory. In contrast to object detection in the general image domain, chart element detection relies heavily on context information as charts are highly structured data visualization formats. To address this, we propose a novel method CACHED, which stands for Context-Aware Chart Element Detection, by integrating a local-global context fusion module consisting of visual context enhancement and positional context encoding with the Cascade R-CNN framework. To improve the generalization of our method for broader applicability, we refine the existing chart element categorization and standardized 18 classes for chart basic elements, excluding plot elements. Our CACHED method, with the updated category of chart elements, achieves state-of-the-art performance in our experiments, underscoring the importance of context in chart element detection. Extending our method to the bar plot detection task, we obtain the best result on the PMC test dataset.