Finiding Peaks and plotting in 3D

Hello everyone I have two problems with my project first is: 1) how i can plot tx, ty and tz? 2) I have to find peaks (I wrote program which check values in matrix 49x49. if "o" value is greater 8 times than all numbers around then that's the peak. Loop unfortunately doesn't work and I can't find a mistake

    1 2 3
47  x x x
48  x o x
49  x x x

My program

%generate a reasonable terrain from a function
tx =linspace(-30, 30, 49*49);
ty = tx;
[xx, yy] = meshgrid(tx, ty);
tz = % I didnt paste it here because data from .m file contain 2401 values. 

figure(1);
mesh(tx, ty, tz);

figure(2);
contour(xx, yy, tz);

for row = 2:48
for column = 2:48
p_c = 0;
t_c = 0;
for n = row - 1:row + 1
for m = column - 1:column + 1

if(tz(row, column) > tz(n, m))
p_c = p_c+1;
elseif(tz(row, column) < tz(n, m))
t_c = t_c+1;
end
end
end
if(p_c > 7)
disp( sprintf( 'Peak found at row %d column %d', row, column));
elseif(t_c > 7)
disp( sprintf( 'Trough found at row %d column %d', row, column));
end
end
end

[Image Analyst]

What does "greater 8 times than all numbers around" mean? Would that mean that the center is greater than the mean of the 8 surrounding values? Or greater than any of the 8 surrounding values? Or the difference between the center and the mean of the 8 surrounding values is 8 times greater than the standard deviation of the surrounding values?

ANd you probably don't need a loop. You can probably use conv2 and thresholding, or a function in the Image Processing Toolbox, like imdilate(), if you have it.

[Kondiik]

if center is greater than all surrounding values then its peak. The problem is our teacher wants us suffer :) we have to find peaks without using function he didn't teach yet.

1. Reads the elevation data from the data file data.m. (A total of 49 × 49 elevation points are taken from a −30 to +30 meters by −30 to +30 meters rectangular terrain.)

2. Generates a 3D surface mesh plot as well as a contour plot of the elevation data

3. Prints the location of peaks in the grid.

[Image Analyst]

Can you use imreagionalmax() in the Image Processing Toolbox?

[Kondiik]

no I have to find it almost "by hand". That's the problem. Last chapter was loops and this one is 3D plotting so he wants use us bothchapters, not ready function in matlab. That's is why I thing many ppl will fail this class.

Answer 1

[A Jenkins]

I ran your code and it works fine for me as long as I change the first line from

tx =linspace(-30, 30, 49*49);

to

tx =linspace(-30, 30, 49);

_____________________________________________________________________

I get the following:

Trough found at row 12 column 27
Peak found at row 20 column 21
Peak found at row 25 column 35
Trough found at row 27 column 14
Trough found at row 28 column 27
Peak found at row 38 column 25

These match what I see in the figures generated.

_____________________________________________________________________

(Though you might want to print out the values of tx(row) and ty(column) instead of just row and column, so it matches the indicies on your plots.)

Trough found at x=-16.250 y=2.500
Peak found at x=-6.250 y=-5.000
Peak found at x=0.000 y=12.500
Trough found at x=2.500 y=-13.750
Trough found at x=3.750 y=2.500
Peak found at x=16.250 y=0.000