1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
|
% function [v,s,d] = firstncut(base_name,rec_num)
% Input:
% base_name = image name
% rec_num = parameter record number
% Output:
% v = eigenvectors
% s = eigenvalues
% d = normalization matrix d = 1/sqrt(rowsum(abs(a)))
% Convert Jianbo Shi's Ncut Ccode results from images to matlab matrices.
% Stella X. Yu, 2000.
function [v,s,d] = firstncut(base_name,rec_num);
if nargin<2 | isempty(rec_num),
rec_num = 1;
end
cur_dir = pwd;
globalenvar;
cd(IMAGE_DIR);
cd(base_name);
feval([base_name,'_par']);
j = length(p);
if rec_num>j,
disp(sprintf('parameter record number %d out of range %d, check %s!',rec_num,j,[base_name,'_par.m']));
Qlabel = [];
v = [];
s = [];
ev_info = [];
return;
end
nv = p(rec_num).num_eigvecs;
no_rep = (p(rec_num).offset<1e-6);
% read the image
cm=sprintf('I = readppm(''%s.ppm'');',base_name);
eval(cm);
% read eigenvectors
base_name_hist = sprintf('%s_%d_IC',base_name,rec_num);
if no_rep,
[v,ev_info] = read_ev_pgm(base_name_hist,1,1,nv);
else
[v,ev_info] = read_ev_pgm2(base_name_hist,1,1,nv);
end
s = ev_info(4,:)';
% read the normalization matrix
d = readpfmc(sprintf('%s_%d_D_IC.pfm',base_name,rec_num));
cd(cur_dir);
% D^(1/2)
dd = (1./(d(:)+eps));
% recover real eigenvectors
for j = 1:nv-no_rep,
vmin = ev_info(1,j);
vmax = ev_info(2,j);
y = v(:,:,j).*((vmax - vmin)/256) + vmin;
%validity check: x = D^(1/2)y should be normalized
x = norm(y(:).*dd);
v(:,:,j) = y./x;
end
dispimg(cat(3,mean(I,3),v),[],[{'image'};cellstr(num2str(s,3))]);
|
