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
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
|
//
// Created by Keuin on 2022/4/11.
//
#ifndef RT_BITMAP_H
#define RT_BITMAP_H
#include <vector>
#include <cstdint>
#include <ostream>
#include <cassert>
#include <vector>
#include "bitfont.h"
#include "vec.h"
#define COLORMIX_OVERFLOW_CHECK
enum text_policy {
hard = 0, // cut off overflown parts
newline = 1, // write overflown characters to new line
};
//// T is some unsigned integer
//template<typename T>
//T saturate_add(T a, T b) {
// T c = a + b;
// if (a > c || b > c) {
// c = (1ULL << (sizeof(T) * 8U)) - 1ULL;
// }
// return c;
//}
// T is some unsigned integer, do not use float point types!
template<typename T>
struct pixel {
T r, g, b;
static constexpr auto mod = (1ULL << (sizeof(T) * 8U)) - 1ULL; // FIXME
bool operator==(const pixel<T> &other) const {
return r == other.r && g == other.g && b == other.b;
}
/**
* Create a pixel of given depth, from normalized color values.
* r, g, b must in range [0, 1].
* For example: Set color depth to 8bit, for normalized color (1, 0.5, 0.25), we get: (255, 127, 63).
*/
static inline pixel<T> from_normalized(double r, double g, double b) {
return pixel<T>{.r = (T) (mod * r), .g = (T) (mod * g), .b = (T) (mod * b)};
}
// v3d must be a normalized vector
static inline pixel<T> from_normalized(const vec3d &v3d) {
return from_normalized(v3d.x / 2.0 + 0.5, v3d.y / 2.0 + 0.5, v3d.z / 2.0 + 0.5);
}
// Convert pixels between different color depths.
template<typename U>
static inline pixel<T> from(const pixel<U> &orig) {
return from_normalized(
1.0 * orig.r / pixel<U>::max_value(),
1.0 * orig.g / pixel<U>::max_value(),
1.0 * orig.b / pixel<U>::max_value()
);
}
static inline pixel<T> black() {
return pixel<T>{(T) 0, (T) 0, (T) 0};
}
static inline pixel<T> white() {
return pixel<T>{(T) mod, (T) mod, (T) mod}; // FIXME float-point values
}
static inline T max_value() {
return mod; // FIXME
}
pixel<T> gamma2() const {
const auto max = max_value();
if (sizeof(T) <= 2) {
// 26% faster than using double
const auto r_ = sqrtf((float) (this->r) / (float) max);
const auto g_ = sqrtf((float) (this->g) / (float) max);
const auto b_ = sqrtf((float) (this->b) / (float) max);
return pixel<T>::from_normalized(r_, g_, b_);
} else {
const auto r_ = sqrt(1.0 * this->r / max);
const auto g_ = sqrt(1.0 * this->g / max);
const auto b_ = sqrt(1.0 * this->b / max);
return pixel<T>::from_normalized(r_, g_, b_);
}
}
};
//template<>
//pixel<uint16_t> pixel<uint16_t>::gamma2() const {
// // 26% faster than using double
// const auto max = max_value();
// const auto r_ = sqrtf((float) (this->r) / (float) (max));
// const auto g_ = sqrtf((float) (this->g) / (float) (max));
// const auto b_ = sqrtf((float) (this->b) / (float) (max));
// return pixel<uint16_t>::from_normalized(r_, g_, b_);
//}
template<
typename T,
typename S,
typename = typename std::enable_if<std::is_arithmetic<S>::value, S>::type
>
pixel<T> operator*(const pixel<T> &pixel, S scale) {
return ::pixel < T > {.r=(T) (pixel.r * scale), .g=(T) (pixel.g * scale), .b=(T) (pixel.b * scale)};
}
template<
typename T,
typename S,
typename = typename std::enable_if<std::is_arithmetic<S>::value, S>::type
>
pixel<T> operator*(S scale, const pixel <T> &pixel) {
return ::pixel < T > {.r=(T) (pixel.r * scale), .g=(T) (pixel.g * scale), .b=(T) (pixel.b * scale)};
}
template<typename S, typename T>
pixel<T> operator*(const vec3<S> &scale, const pixel <T> &pixel) {
return ::pixel < T > {.r=(T) (pixel.r * scale.x), .g=(T) (pixel.g * scale.y), .b=(T) (pixel.b * scale.z)};
}
// Mix two colors a and b. Returns a*u + b*v
template<typename T>
inline pixel<T> mix(const pixel<T> &a, const pixel<T> &b, double u, double v) {
assert(u >= 0);
assert(v >= 0);
assert(u <= 1);
assert(v <= 1);
assert(u + v <= 1);
pixel<T> c{0, 0, 0};
c.r = (T) (u * a.r + v * b.r);
c.g = (T) (u * a.g + v * b.g);
c.b = (T) (u * a.b + v * b.b);
return c;
}
// 8 bit pixel
using pixel8b = pixel<uint8_t>;
template<typename T>
class bitmap {
unsigned width, height;
std::vector<pixel<T>> content; // pixels scanned by rows, from top to bottom
pixel<T> &image(unsigned x, unsigned y) {
assert(x < width);
assert(y < height);
return content[x + y * width];
}
pixel<T> &image(unsigned x, unsigned y) const {
assert(x < width);
assert(y < height);
return content[x + y * width];
}
public:
bitmap() = delete;
bitmap(unsigned int width, unsigned int height) :
width(width), height(height), content{width * height, pixel<T>{}} {}
bitmap(unsigned int width, unsigned int height, std::vector<pixel<T>> &&data) :
width(width), height(height), content{data} {}
static bitmap<T> average(const std::vector<bitmap<T>> &images) {
using Acc = typename std::conditional<
(sizeof(T) <= 1),
uint_fast16_t,
typename std::conditional<
(sizeof(T) <= 2),
uint_fast32_t,
typename std::conditional<
(sizeof(T) <= 4),
uint_fast64_t,
uintmax_t
>::type
>::type
>::type; // pick the smallest suitable type for accumulator
static_assert(sizeof(Acc) > sizeof(T), "accumulator may overflow");
assert(!images.empty());
bitmap<T> result{images[0].width, images[0].height};
const auto m = images.size();
const auto n = images[0].content.size();
Acc acc_r, acc_g, acc_b;
for (size_t i = 0; i < n; ++i) {
acc_r = 0;
acc_g = 0;
acc_b = 0;
for (size_t j = 0; j < m; ++j) {
acc_r += images[j].content[i].r;
acc_g += images[j].content[i].g;
acc_b += images[j].content[i].b;
}
result.content[i] = pixel<T>{(T) (1.0 * acc_r / m), (T) (1.0 * acc_g / m), (T) (1.0 * acc_b / m)};
}
return result;
}
void set(unsigned x, unsigned y, const pixel<T> &pixel) {
image(x, y) = pixel;
}
pixel<T> get(unsigned x, unsigned y) const {
return image(x, y);
}
const pixel<T> &operator[](size_t loc) const {
assert(loc < width * height);
return content[loc];
}
pixel<T> &operator[](size_t loc) {
assert(loc < width * height);
return content[loc];
}
// Do not use float-point pixels, or this routine will break.
void write_plain_ppm(std::ostream &out) const;
// Draw text on the image. Supports limited visual ASCII characters.
void print(const std::string &s, const pixel<T> &color,
unsigned x, unsigned y, text_policy policy, unsigned scale = 1, double alpha = 1.0);
bool normalized() const {
return false;
// TODO return true for float-point pixels
}
std::pair<unsigned, unsigned> shape() const {
return std::pair<unsigned, unsigned>{width, height};
}
// U: original color depth, T: desired color depth
template<typename U>
static bitmap<T> from(const bitmap<U> &src) {
const auto shape = src.shape();
const size_t sz = shape.first * shape.second;
bitmap<T> out{shape.first, shape.second};
for (size_t i = 0; i < sz; ++i) {
out[i] = pixel<T>::from(src[i]);
}
return out;
}
bitmap<T> gamma2() const {
std::vector<pixel<T>> out;
out.reserve(content.size());
for (const auto &pix: content) {
out.push_back(pix.gamma2());
}
return bitmap<T>{width, height, std::move(out)};
}
};
template<typename T>
void bitmap<T>::write_plain_ppm(std::ostream &out) const {
const auto depth = (1ULL << (sizeof(T) * 8)) - 1ULL; // color depth of pixels
out << "P3\n" // file header
<< width << ' ' << height << '\n' // image width and height
<< depth << '\n'; // normalized max color depth (e.g. 255 for 8bit image)
for (const auto &pixel: content) {
out << (unsigned long long) pixel.r << ' '
<< (unsigned long long) pixel.g << ' '
<< (unsigned long long) pixel.b << '\n';
}
}
template<typename T>
void bitmap<T>::print(const std::string &s, const pixel<T> &color,
unsigned x, unsigned y, text_policy policy, unsigned scale, double alpha) {
assert(alpha >= 0);
assert(alpha <= 1);
const unsigned char_w = 8 * scale, char_h = 13 * scale; // char width and height
size_t n = 0; // written characters
for (const auto &c: s) {
unsigned int idx = c - 32U;
if (idx >= 95) idx = 1; // replace invisible chars with '!'
unsigned char_x, char_y;
const unsigned spacing = ((n > 1) ? ((n - 1) * scale) : 0); // total spacing between characters
if (policy == text_policy::hard) {
char_x = x + n * char_w + spacing;
char_y = y;
} else if (policy == text_policy::newline) {
const auto newlines = (n * char_w + spacing + char_w - 1) / (width - x);
char_x = (x + n * char_w + spacing) % (width - x);
char_y = y + newlines * char_h + ((newlines > 0) ? ((newlines - 1) * scale) : 0);
} else {
abort(); // unknown policy
}
// char size is 13x8, stored line by line, from the bottom line to the top line
// every line is represented in a single byte
// in every byte, from MSB to LSB, the pixel goes from left to right
for (unsigned i = 0; i < char_w; ++i) {
for (unsigned j = 0; j < char_h; ++j) {
// downscale (i, j) to pos on char bitmap i -> i/scale, j -> j/scale
if (rasters[idx][12 - j / scale] & (1U << (7U - i / scale))) {
const auto pen_x = char_x + i, pen_y = char_y + j;
if (pen_x >= width || pen_y >= height) continue;
image(pen_x, pen_y) = mix(image(pen_x, pen_y), color, 1.0 - alpha, alpha);
}
}
}
++n;
}
}
using bitmap8b = bitmap<uint8_t>;
#endif //RT_BITMAP_H
|