Automate the Process of Image Recognizing a Scatter Plot: an Application of Non-parametric Cluster Analysis in Capturing Data from Graphical Output

A fundamental method of pharmaceutical research is to compare the indication between a new drug and an existing competitor drug. Sometimes, the original data of the competitor drug may not always be available. What a researcher can access is only a plot from a published paper. To make the comparison, the researcher has to capture tabular data from this legacy graphical output.