My attempt at creating 3x3 kernel to analyse surround pixels however …

…this had several drawback the edges were not included and speed was very slow.

convulution.m

@@ -0,0 +1,43 @@
+%  misalignment
+%
+% Date:18/2/11
+%
+% 
+%% Clear workspace
+
+clear all
+close all
+clc
+
+%% Image Upload
+
+%image test.tif was saved in a directory found in search path
+image_1='test.tif';
+
+% Reads image data
+h=imread(image_1);
+h2=h;
+[y,x]=find(h>=0);
+
+% These find the min and max co-ordinates to allow the creation of a
+% rectangular array needed for creation of plot
+max_x=max(x);
+max_x=max_x-1;
+min_x=min(x);
+min_x=min_x+1;
+
+% Below ensures that no edges are detected in h2
+max_y=max(y);
+max_y=max_y-1;
+min_y=min(y);
+min_y=min_y+1;
+
+for x=min_x:max_x;
+    for y=min_y:max_y;
+        moving_array=h2(y-1:y+1,x-1:x+1);% 3x3 kernel
+        average_1=mean(moving_array);% 3 values one for each row
+        h2(y,x)=round(mean(average_1));% takes all averages and rounds them to an integer and places it x y of current loop
+    end
+end
+
+% Program works but is very slow