A Two-stage Framework for Compound Figure Separation

TL;DR

This paper introduces a two-stage framework for separating complex compound figures into individual subfigures, improving detection precision and aiding information retrieval from scientific images.

Contribution

The paper presents a novel two-stage method that detects subfigure labels first and then uses this information to accurately segment subfigures, enhancing previous approaches.

Findings

Improves detection precision by 9%.

Effectively preserves subfigure-caption associations.

Validated through extensive experiments.

Abstract

Scientific literature contains large volumes of complex, unstructured figures that are compound in nature (i.e. composed of multiple images, graphs, and drawings). Separation of these compound figures is critical for information retrieval from these figures. In this paper, we propose a new strategy for compound figure separation, which decomposes the compound figures into constituent subfigures while preserving the association between the subfigures and their respective caption components. We propose a two-stage framework to address the proposed compound figure separation problem. In particular, the subfigure label detection module detects all subfigure labels in the first stage. Then, in the subfigure detection module, the detected subfigure labels help to detect the subfigures by optimizing the feature selection process and providing the global layout information as extra features. Extensive experiments are conducted to validate the effectiveness and superiority of the proposed framework, which improves the detection precision by 9%.