Dear community, just wondering if it is possible to extract values from a data sheet in a JPG or PDF format. For instance I have a certain data sheet graph (i.e. battery discharge voltage vs time) saved in a PDF or JPG format. Is there any way to convert the image into actual values? i.e. converting the picture of the graph into the values actually assumed by the graph? Thank you
A harder method would be take the matrix of points from ImageData, figure out the
RGB values for the curves, and use Position to get the pixel coordinates of points corresponding to each curve. You might then want to do some data smoothing to get one thin curve.
Select a few points on the smoother curve and to interpolation. Then add the sharp part.
Adding to Bruce’s suggestion, I’d try doing a few Erosion[] steps to remove the letters (which I think might have same rgb values) and the dashed grid lines. Subsequently, You might also try a MorphologicalComponents and see if you can grab the connected pixels. Average the Y pixel values at each common x pixel value, and then use Interpolation.
Take a look at this post on mathematica.stackexchange.com, where a scatter plot digitizer is shown.