Perceptual Interpretation of Ink Annotations on Line Charts

Summary:  When giving a presentation on a chart like the one above, one would typically draw an arrow to annotate the visually salient parts of the graph that attention needs to be drawn to.  However, what would be optimal would be to highlight a certain part of the graph as well but this kind of detail would be impossible to implement on the fly during a presentation.  What the researchers behind this paper have done is utilize Gestalt's laws of visual perception to take an educated guess at what the presenter wants to highlight and does the tedious work for them.  In this case it would find the peak that the arrow is pointing to and high light that peak or find the decline that was marked and highlight that portion of the line.

This technology would have several implications or possible extensions. If a corresponding data table is paired with a chart, finding what part of the graph is being highlighted could also highlight the corresponding data within the data table.

The way this was approached was to divide charts into maxima, minima, and slopes.   They classified strokes as either parallel or roughly orthogonal. They assumed parallel strokes would generally highlight slopes and orthogonal strokes would generally highlight entire peaks or valleys.

Discussion:

I feel like the concept as it is is very limited in application.  As it stands, it only works with line graphs and the researchers themselves said that they needed to conduct more research to see if their research was effective.  I would like to see the concept extended to apply to general annotations because I feel like this would aid in school lectures especially when having to draw attention to minute details with in complex and intricate schematic such as a certain part within a circuit diagram.