Panda3D
Loading...
Searching...
No Matches
pnm-image-filter-sparse-core.cxx
Go to the documentation of this file.
1/**
2 * PANDA 3D SOFTWARE
3 * Copyright (c) Carnegie Mellon University. All rights reserved.
4 *
5 * All use of this software is subject to the terms of the revised BSD
6 * license. You should have received a copy of this license along
7 * with this source code in a file named "LICENSE."
8 *
9 * @file pnm-image-filter-sparse-core.cxx
10 * @author drose
11 * @date 2013-01-25
12 */
13
14// We map X and Y to A and B, because we might change our minds about which is
15// dominant, and we map getset functions for the channel in question to
16// GETVALSETVAL.
17
18
19static void
20FUNCTION_NAME(IMAGETYPE &dest, const IMAGETYPE &source,
21 float width, FilterFunction *make_filter, int channel) {
22 if (!dest.is_valid() || !source.is_valid()) {
23 return;
24 }
25
26 // First, set up a 2-d column-major matrix of StoreTypes, big enough to hold
27 // the image xelvals scaled in the A direction only. This will hold the
28 // adjusted xel data from our first pass.
29
30 StoreType **matrix = (StoreType **)PANDA_MALLOC_ARRAY(dest.ASIZE() * sizeof(StoreType *));
31 StoreType **matrix_weight = (StoreType **)PANDA_MALLOC_ARRAY(dest.ASIZE() * sizeof(StoreType *));
32
33 int a, b;
34
35 for (a=0; a<dest.ASIZE(); a++) {
36 matrix[a] = (StoreType *)PANDA_MALLOC_ARRAY(source.BSIZE() * sizeof(StoreType));
37 matrix_weight[a] = (StoreType *)PANDA_MALLOC_ARRAY(source.BSIZE() * sizeof(StoreType));
38 }
39
40 // First, scale the image in the A direction.
41 float scale;
42 StoreType *temp_source, *temp_source_weight, *temp_dest, *temp_dest_weight;
43
44 scale = (float)dest.ASIZE() / (float)source.ASIZE();
45 temp_source = (StoreType *)PANDA_MALLOC_ARRAY(source.ASIZE() * sizeof(StoreType));
46 temp_source_weight = (StoreType *)PANDA_MALLOC_ARRAY(source.ASIZE() * sizeof(StoreType));
47 temp_dest = (StoreType *)PANDA_MALLOC_ARRAY(dest.ASIZE() * sizeof(StoreType));
48 temp_dest_weight = (StoreType *)PANDA_MALLOC_ARRAY(dest.ASIZE() * sizeof(StoreType));
49
50 WorkType *filter;
51 float filter_width;
52
53 make_filter(scale, width, filter, filter_width);
54
55 for (b = 0; b < source.BSIZE(); b++) {
56 memset(temp_source_weight, 0, source.ASIZE() * sizeof(StoreType));
57 for (a = 0; a < source.ASIZE(); a++) {
58 if (source.HASVAL(a, b)) {
59 temp_source[a] = (StoreType)(source_max * source.GETVAL(a, b, channel));
60 temp_source_weight[a] = filter_max;
61 }
62 }
63
64 filter_sparse_row(temp_dest, temp_dest_weight, dest.ASIZE(),
65 temp_source, temp_source_weight, source.ASIZE(),
66 scale,
67 filter, filter_width);
68
69 for (a = 0; a < dest.ASIZE(); a++) {
70 matrix[a][b] = temp_dest[a];
71 matrix_weight[a][b] = temp_dest_weight[a];
72 }
73 }
74
75 PANDA_FREE_ARRAY(temp_source);
76 PANDA_FREE_ARRAY(temp_source_weight);
77 PANDA_FREE_ARRAY(temp_dest);
78 PANDA_FREE_ARRAY(temp_dest_weight);
79 PANDA_FREE_ARRAY(filter);
80
81 // Now, scale the image in the B direction.
82 scale = (float)dest.BSIZE() / (float)source.BSIZE();
83 temp_dest = (StoreType *)PANDA_MALLOC_ARRAY(dest.BSIZE() * sizeof(StoreType));
84 temp_dest_weight = (StoreType *)PANDA_MALLOC_ARRAY(dest.BSIZE() * sizeof(StoreType));
85
86 make_filter(scale, width, filter, filter_width);
87
88 for (a = 0; a < dest.ASIZE(); a++) {
89 filter_sparse_row(temp_dest, temp_dest_weight, dest.BSIZE(),
90 matrix[a], matrix_weight[a], source.BSIZE(),
91 scale,
92 filter, filter_width);
93
94 for (b = 0; b < dest.BSIZE(); b++) {
95 if (temp_dest_weight[b] != 0) {
96 dest.SETVAL(a, b, channel, (float)temp_dest[b]/(float)source_max);
97 }
98 }
99 }
100
101 PANDA_FREE_ARRAY(temp_dest);
102 PANDA_FREE_ARRAY(temp_dest_weight);
103 PANDA_FREE_ARRAY(filter);
104
105 // Now, clean up our temp matrix and go home!
106
107 for (a = 0; a < dest.ASIZE(); a++) {
108 PANDA_FREE_ARRAY(matrix[a]);
109 PANDA_FREE_ARRAY(matrix_weight[a]);
110 }
111 PANDA_FREE_ARRAY(matrix);
112 PANDA_FREE_ARRAY(matrix_weight);
113}