/
rgb2ycbcr_double.m
137 lines (116 loc) · 3.69 KB
/
rgb2ycbcr_double.m
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
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
function [varargout] = rgb2ycbcr_double(varargin)
% RGB2YCBCR_DOUBLE
% Convert image from RGB space into YCbCr space
% SYNTAX
% [ycbcr] = rgb2ycbcr_double(rgb);
% [y, cb, cr] = rgb2ycbcr_double(r,g,b);
% [...] = rgb2ycbcr_double(...,'Flag',...);
% DESCRIPTION
% Takes 'rgb', an (M,N,3) RGB double precision image, and converts
% 'rgb' into an YCbCr image, 'ycbcr'. Alternately, each image plane may
% be passed separately, in 'r', 'g' and 'b' input arguments.
% By default, the return image format will be the same as the input image
% format (i.e., one variable vs each plane separately)
%
% Nominal input Y values are on [16,235] and nominal input Cb and Cr values
% are on [16,240]. RGB values are on [0,255]. This routine does not
% round final RGB values to the nearest integer.
%
% The following optional arguments may be specified:
%
% '128' Return Cb and Cr values on a range from -128 to 127.
% By default, Cb and Cr range from [0..255]
% 'img' Input variable 'rgb' is in 'img' format, as per imread
% (i.e., [0..1] instead of [0..255]
% 'ycbcr' Return matrix size (M,N,3), where (:,:,1) contains the Y
% plane, (:,:,2) Cb, and (:,:,3) Cr.
% 'y_cb_cr' Return Y, Cb and Cr planes in three separate variables,
% each size (M,N).
%
%
% Reference:
% Charles Poynton ColorFAQ.pdf (page 15), available from www.poynton.com.
%
is_128 = 0;
is_img = 0;
want_y_cb_cr = 0;
want_ycbcr = 0;
%
num_images = 0;
num_flags = 0;
for cnt=1:nargin
if isnumeric(varargin{cnt})
num_images = num_images + 1;
elseif ischar(varargin{cnt})
num_flags = num_flags + 1;
else
error('input arguments to "rgb2ycbcr_double" are unrecognizable');
end
end
if num_images == 1
red = varargin{1}(:,:,1);
green = varargin{1}(:,:,2);
blue = varargin{1}(:,:,3);
want_ycbcr = 1;
elseif num_images == 3
red = varargin{1};
green = varargin{2};
blue = varargin{3};
want_y_cb_cr = 1;
else
error('unexpected number of image matrices input to rgb2ycbcr_double');
end
% parse flags
for cnt = num_images+1:num_images+num_flags
if strcmp(varargin{cnt},'128')
is_128 = 1;
elseif strcmp(varargin{cnt},'img')
is_img = 1;
red = red * 256;
green = green * 256;
blue = blue * 256;
elseif strcmp(varargin{cnt},'ycbcr')
want_y_cb_cr = 0;
want_ycbcr = 1;
elseif strcmp(varargin{cnt},'y_cb_cr')
want_y_cb_cr = 1;
want_ycbcr = 0;
else
error('input flag not recognized');
end
end
%---------------------------------------
% This is inefficient code for debugging
if 0
% Transformation for each pixel is given by:
% [Y Cb Cr]' = a0 + a1 * [R G B]'; RGB on [0, 255]
a0 = [16;128;128]; % Offset
a1 = [65.481 128.553 24.966; -37.797 -74.203 112; 112 -93.786 -18.214]/255; % Matrix
ycbcr = repmat(a0,1,nr*nc) + a1*reshape(rgb,nr*nc,np)';
ycbcr = reshape(ycbcr',nr,nc,np);
% Clip at 0 and 255
ycbcr1 = max(0, min(ycbcr, 255));
end
%---------------------------------------
% convert RGB to YCbCr
y = 16.0 + ((65.481/255.0)*red) + ((128.553/255.0)*green) + ((24.966/255.0)*blue);
cb = 128.0 + ((-37.797/255.0)*red) + ((-74.203/255.0)*green) + ((112.0/255.0)*blue);
cr = 128.0 + ((112.0/255.0)*red) + ((-93.786/255.0)*green) + ((-18.214/255.0)*blue);
% shift, if wanted
if is_128
cb = cb - 128;
cr = cr - 128;
end
% format return variables
if want_y_cb_cr
varargout{1} = y;
varargout{2} = cb;
varargout{3} = cr;
else
[row,col] = size(red);
ycbcr = nan(row,col,3);
ycbcr(:,:,1) = y;
ycbcr(:,:,2) = cb;
ycbcr(:,:,3) = cr;
varargout{1} = ycbcr;
end