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
|
#include <chrono>
#include <cstdio>
#include "compositing.h"
#ifndef NO_AUTO_SCHEDULE
#include "compositing_auto_schedule.h"
#endif
#include "HalideBuffer.h"
#include "halide_benchmark.h"
#include "halide_image_io.h"
using namespace Halide::Runtime;
using namespace Halide::Tools;
// MSVC doesn't define this by default
#ifndef M_PI
#define M_PI 3.14159265358979310000
#endif
int main(int argc, char **argv) {
if (argc < 4) {
printf("Usage: ./process input.png timing_iterations output.png\n"
"e.g.: ./process input.png 10 output.png\n");
return 1;
}
Buffer<uint8_t, 3> input = load_and_convert_image(argv[1]);
Buffer<uint8_t, 3> output(input.width(), input.height(), 4);
int timing = atoi(argv[2]);
// Make a ring of colored blobs to composite over the input, each using a different blend mode
Buffer<uint8_t, 3> blobs[5];
int op_codes[] = {4, 3, 2, 1, 0}; // These op codes specify the blend mode to use for each layer.
Buffer<int, 1> ops(op_codes);
for (int i = 0; i < 5; i++) {
blobs[i] = Buffer<uint8_t, 3>::make_with_shape_of(input);
blobs[i].fill(255);
for (int y = 0; y < blobs[i].height(); y++) {
for (int x = 0; x < blobs[i].width(); x++) {
int cx = cos(i * 2 * M_PI / 5) * 300 + input.width() / 2;
int cy = sin(i * 2 * M_PI / 5) * 300 + input.height() / 2;
int alpha = std::min(255, std::min(std::max(0, 500 - std::abs(x - cx)), std::max(0, 500 - std::abs(y - cy))));
blobs[i](x, y, 3) = alpha;
blobs[i](x, y, 0) = 255;
blobs[i](x, y, 1) = (255 / 3) * i;
blobs[i](x, y, 2) = 255 - blobs[i](x, y, 1);
}
}
}
compositing(input, blobs[0], blobs[1], blobs[2], blobs[3], blobs[4], ops, output);
// Timing code
// Manually-tuned version
double best_manual = benchmark(timing, 1, [&]() {
compositing(input, blobs[0], blobs[1], blobs[2], blobs[3], blobs[4], ops, output);
output.device_sync();
});
printf("Manually-tuned time: %gms\n", best_manual * 1e3);
#ifndef NO_AUTO_SCHEDULE
// Auto-scheduled version
double best_auto = benchmark(timing, 1, [&]() {
compositing_auto_schedule(input, blobs[0], blobs[1], blobs[2], blobs[3], blobs[4], ops, output);
output.device_sync();
});
printf("Auto-scheduled time: %gms\n", best_auto * 1e3);
#endif
convert_and_save_image(output, argv[3]);
printf("Success!\n");
return 0;
}
|
