Given the extracted figures and captions, we first assign them a label corresponding to their structural position (e.g., left column, right column, and double column).
We solved the field equation with a tensor Green's function technique, and extracted figures of merit describing the influence of the vacuum pipe on the radiation pulse as a function of the problem parameters.
Given the extracted figure-caption pairs, we first identify the number of panels from the caption and, separately, identify the number of panels from the image.
Specifically, we evaluate the position of each extracted figure with respect to a pre-computed rectangle bounding the articles text, and exclude the figures that are outside the rectangle.
Moreover, the corresponding 3D points belonging to the same types of building elements could be extracted (Figure 7).
For fine mode aerosol, three PCs were extracted (Figure 14).
For coarse mode aerosol, four PCs were extracted (Figure 14).
In the Yale Image Finder, Xu et al. [24] developed algorithms to extract text strings from figures and provided a figure searching tool that queries over the extracted figure text as well as caption text.
All relevant secondary care records were identified and data extracted (figure 1).
We chose to use this dataset because we are particularly interested in extracting figures related to PPI assertions, and also the accompanying textual description of the methods, which is usually found within subcaptions or text embedded in panels.
