如何在Matplotlib图中提取特定点？ （蟒蛇）

Question

From an image, I have computed a mean value over each column and created a graph. This is the output (without letters and marked points).

Most graphs have a behavior like graphs marked with blue, green and red color. From those graphs, I would like to extract point C . But sometimes A is extracted instead of C. Therefore I would like to find an unambiguous solution.

Idea:

• The interval from A to B is always shorter than C to D.

Problem:

• I don't know how to extract point A, B, C.
• Not all graphs are the same (example: light blue line). How to distinguish between them?

Edit #1

Answer 1

The light blue curve being completely different, I would start by detecting it. Fir that I would use three time points (50, 700 and 1400). You see that the values |V50-V700|, |V700-V1400| are completely different for the two types of curves (Blue green red versus light blue).

Then you can detect the A, B and C points. They are characterized by the important and sudden variations (so gradient). I would start by using a median filter in order to denoise a little bit your signals, and then I would compute the signal gradient. A, B and C are going to be the three points with the biggest gradients.