get rid of rugged edges plotted with trisurf

I am using trisurf to plot a set of scattered data (that happens to be cartesian coordinates of a spherical surface whose equation is unknown). Further, there is a parameter Pindex for every (x,y,z) according to which colormap is used. However, the region of one particular color is having severely rugged edges. The link shows a part of image:

http://postimage.org/image/nh4l31ybr/

I am using this code for plotting:

tri=delaunay(x,y); trisurf(tri,x,y,z,Pindex);

I tried using griddata and gridfit but they are not giving proper results. Does anyone know what modification I need to do in my code to get smooth edges of the regions?

[Ryan]

Could you do a convolution with some gaussian to smooth it? I believe the conv2() filter could help you.

[Ryan]

I also found this through google, I believe it will help you http://stackoverflow.com/questions/6330790/3d-gaussian-filter-in-matlab.

Answer 1

[Saurabh Singh]

Thanks Anton and others for the reply.

Actually I'm already using shading interp that I forgot to mention in my question.

Further, based on suggestions, I used imfilter command and tried Gaussian and few other filters on my plot. But the shape of plot (that needs to be spherical) is not being preserved.

1) Is there any image processing tool (like imfilter) that I can use in my case? 2) Is there any straight way without going to image processing to smoothen edges of few parts of the plot?

Answer 2

[Patrick Kalita]

Try changing the figure's renderer to ZBuffer:

set(gcf, 'Renderer', 'zbuffer');

Answer 3

[Saurabh Singh]

Thanks Kalita for the try...but isn't working as well....

Can anyone have an idea if any type of filter can do the job or may be any other surface plot command (that can plot from scattered datapoints (x,y,z) )