1. Laboratory of Image Processing Pier Luigi Mazzeo
July 23, 2014 1

2. Filtering

3. Neighbourhood processing

4. Mean

5. Filtering in MatLab

6. filter2

7. filter2 (cont.)

8. fspecial (low pass filters)
Produce an Average filtering image using 9x9 filter and a 25x25 filter

9. fspecial (low pass filters)
>> hsize = 10;
>> sigma = 5;
>> h = fspecial(‘gaussian’ hsize, sigma);
>> mesh(h);
>> imagesc(h);
>> outim = imfilter(c, h);
>> imshow(outim);
In matlab, conv2 does convolution, filter2 does correlation. Imfilter does either if specified, correlation by default (‘conv’, ‘corr’ option)

10. fspecial (low pass filters)
for sigma=1:3:10
h = fspecial('gaussian‘, fsize, sigma);
out = imfilter(c, h);
imshow(out);
pause;
end

11. Noise Periodic noise >> noise = randn(size(c)).*sigma;
>> output = c+ noise;
Periodic noise

12. Cleaning salt and pepper noise

13. fspecial (high pass filters)

14. fspecial (high pass filters)

15. Unsharp masking
>> My = fspecial(‘sobel’);
>> outim = imfilter(double(c), My);
>> imagesc(outim);
>> colormap gray;
Find the image ‘pelicans.tif’ from internet and use u filter

16. Canny edge detector MATLAB: edge(c, ‘canny’); >>help edge
I = imread('circuit.tif');
BW1 = edge(I,'prewitt');
BW2 = edge(I,'canny');
figure, imshow(BW1)
figure, imshow(BW2)