Is there a robust matlab algorithm for spatial smoothing in all directions?

I got a 2-D matrix of size 360X180 that represent geogrpahical locations (lon,lat). The values in the matrix is probability to find something at that geographical location. I want to smooth that matrix to get better plots. Until now I used matlab's filtfilt function to smooth the data, however I realise that besides the fact that it misses the diagonal adjacent values, it also matters if I choose to run it first alongside the lat or the lon dimension. I searched quite a lot but couldn't find any matlab function that can give me a nice spatial data smoothing that is robust, and also that is a bit more sophisticated algorithmically than just averaging the matrix with neighbouring values. If you have any idea of a matlab function that can achieve this I'd be very happy to know!

[Rik]

You mean like a Gaussian blur? You can use a convolution to do that.

You should note that a 'better plot' is a very subjective goal to work towards. What does that mean exactly? And why do you want something sophisticated?

[Walter Roberson]

Are the matrices equally spaced in latitude, and equally spaced in longitude? Or do they correspond more to physical coordinates (for example, spaced in km instead of minutes of a degree) ?