Hello. How algorithm I can use for detect the start point of the fast increasing of my data and the end of this procces?

My data corrupted by some noise, I use the diff function for indicate the start point(labelled by red point on the figure), but find it only with large delay and don't know how to indicate correctly the end of increase (labelled by vertical line on the figure).

Answer 1

[Walter Roberson]

dd = diff(data);
mask = abs(dd) > SomeThreshold;
find(mask, 1, 'first')
find(mask, 1, 'last')

[Zak Kankin]

It works only for well smoothing data. If I use the mooving average filter for this purpose I receive large displace for the peaks.