How can I remove noisy data from my plots

I am doing simulation for kinematic analysis of rover using matlab. But due to discretization of the terrain I am getting some noisy data in my graphs which comes as peaks at the connecting points when I am calculating velocity-ratios. How can I remove them.... I dont want to fit a curve or smoothen the graph..I just want to remove noise... Thanks in advance for any sort of help..

1 Comment

Remove noise, but do not smoothen? Isn't this a contradiction?
Can you define the properties of the peaks mathematically?

Sign in to comment.

 Accepted Answer

There are tons of noise reduction filters. Way too many to list here. Maybe you can get by with a simple median filter. If that doesn't work, then say why not.

More Answers (3)

X(X>some_noise_threshold) = [];
removes stuff greater than the threshold you set for noise. Show us a small set of sample data or a picture of the plot for more details.

2 Comments

Actually there are some peaks coming in plots and I want to reomve them only without smoothing the graphs but cant decide any threshold cause its variable...
Don't know how to attach any image here...
See http://www.mathworks.com/matlabcentral/answers/7924-where-can-i-upload-images-and-files-for-use-on-matlab-answers

Sign in to comment.

Perhaps Brett Shoelson's "deleteoutliers" function may be what you want: http://www.mathworks.com/matlabcentral/fileexchange/3961
Which toolboxes do you have? Do you have the Signal Processing or Image Processing Toolboxes? You could use a smoothing filter such as a box filter with conv(), or sgolay(), or medfilt1(). Then subtract your original image and find the indexes (with the find() function) where the difference exceeds some threshold and replace those indexes in the original data with the smoothed data.

2 Comments

yep i did the same used medfilt1()...
Thanks for the help..
yep i did the same used medfilt1()...
Thanks for the help..

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!