Data Segmentation

Good evening to everyone, I would like to ask some help. I have a data that varies with time and that is sampled at 200 Hz. this data is characterized by some "bursts" (some groups of peaks of my signal ) I would like to know some ways i can segment my datas and in particular isolate the different burst. I would define a burst as a group of peaks in my signals that is distant at least 3 seconds from the next one group of peaks.

Anyone can give me suggestions on how to do this? thanks

Answer 1

[Image Analyst]

Can you supply some sample data? In the meantime, you can use findpeaks if you have the toolbox that it's in. Otherwise threshold and send the result into find() and start scanning for gaps of 3 seconds. It's all a lot easier if you have the Image Processing Toolbox where you can use functions like bwdist() and bwlabel() and regionprops() to find data separated by certain distances and of certain sizes.

Finally, define exactly what you mean by "isolate" since this is somewhat vague and ambiguous. Do you want to know the location of these peaks, or do you want to extract them out into their own submatrix? What exactly does "isolate" mean to you?