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
|
/* Copyright (c) 2008-2022 the MRtrix3 contributors.
*
* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/.
*
* Covered Software is provided under this License on an "as is"
* basis, without warranty of any kind, either expressed, implied, or
* statutory, including, without limitation, warranties that the
* Covered Software is free of defects, merchantable, fit for a
* particular purpose or non-infringing.
* See the Mozilla Public License v. 2.0 for more details.
*
* For more details, see http://www.mrtrix.org/.
*/
#ifndef __math_sinc_h__
#define __math_sinc_h__
#include "math/math.h"
namespace MR
{
namespace Math
{
template <typename T = float> class Sinc
{ NOMEMALIGN
public:
using value_type = T;
Sinc (const size_t w) :
window_size (w),
max_offset_from_kernel_centre ((w-1) / 2),
indices (w),
weights (w),
current_pos (NAN)
{
assert (w % 2);
}
template <class ImageType>
void set (const ImageType& image, const size_t axis, const value_type position) {
if (position == current_pos)
return;
const int kernel_centre = std::round (position);
value_type sum_weights = 0.0;
for (size_t i = 0; i != window_size; ++i) {
const int voxel = kernel_centre - max_offset_from_kernel_centre + i;
if (voxel < 0)
indices[i] = -voxel - 1;
else if (voxel >= image.size (axis))
indices[i] = (2 * int(image.size (axis))) - voxel - 1;
else
indices[i] = voxel;
const value_type offset = position - (value_type)voxel;
const value_type sinc = offset ? std::sin (Math::pi * offset) / (Math::pi * offset) : 1.0;
//const value_type hann_cos_term = Math::pi * offset / (value_type(max_offset_from_kernel_centre) + 0.5);
//const value_type hann_factor = (abs (hann_cos_term) < Math::pi) ? 0.5 * (1.0 + std::cos (hann_cos_term)) : 0.0;
//const value_type this_weight = hann_factor * sinc;
const value_type lanczos_sinc_term = abs (Math::pi * offset / (double(max_offset_from_kernel_centre) + 0.5));
value_type lanczos_factor = 0.0;
if (lanczos_sinc_term < Math::pi) {
if (lanczos_sinc_term)
lanczos_factor = std::sin (lanczos_sinc_term) / lanczos_sinc_term;
else
lanczos_factor = 1.0;
}
const value_type this_weight = lanczos_factor * sinc;
weights[i] = this_weight;
sum_weights += this_weight;
}
const value_type normalisation = 1.0 / sum_weights;
for (size_t i = 0; i != window_size; ++i)
weights[i] *= normalisation;
current_pos = position;
}
size_t index (const size_t i) const { return indices[i]; }
template <class ImageType>
value_type value (ImageType& image, const size_t axis) const {
assert (current_pos != NAN);
const size_t init_pos = image.index(axis);
value_type sum = 0.0;
for (size_t i = 0; i != window_size; ++i) {
image.index(axis) = indices[i];
sum += image.value() * weights[i];
}
image.index(axis) = init_pos;
return sum;
}
template <class Cont>
value_type value (Cont& data) const {
assert (data.size() == window_size);
assert (current_pos != NAN);
value_type sum = 0.0;
for (size_t i = 0; i != window_size; ++i)
sum += data[i] * weights[i];
return sum;
}
value_type value (value_type* data) const {
assert (current_pos != NAN);
value_type sum = 0.0;
for (size_t i = 0; i != window_size; ++i)
sum += data[i] * weights[i];
return sum;
}
private:
const size_t window_size, max_offset_from_kernel_centre;
vector<size_t> indices;
vector<value_type> weights;
value_type current_pos;
};
}
}
#endif
|
