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
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
|
/* pbmtomrf - convert pbm to mrf
* public domain by RJM
*
* Adapted to Netpbm by Bryan Henderson 2003.08.09. Bryan got his copy from
* ftp://ibiblio.org/pub/linux/apps/convert, dated 1998.03.03.
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <limits.h>
#include "pm_c_util.h"
#include "pbm.h"
typedef struct bitOut {
int bitbox;
int bitsleft;
FILE * fileP;
} bitOut;
static void
bit_init(struct bitOut * const bitOutP,
FILE * const ofP) {
bitOutP->bitbox = 0;
bitOutP->bitsleft = 8;
bitOutP->fileP = ofP;
}
static void
bit_output(struct bitOut * const bitOutP,
int const bit) {
--bitOutP->bitsleft;
bitOutP->bitbox |= (bit << bitOutP->bitsleft);
if (bitOutP->bitsleft == 0) {
fputc(bitOutP->bitbox, bitOutP->fileP);
bitOutP->bitbox = 0;
bitOutP->bitsleft = 8;
}
}
static void
bit_flush(struct bitOut * const bitOutP) {
/* there are never 0 bits left outside of bit_output, but
* if 8 bits are left here there's nothing to flush, so
* only do it if bitsleft!=8.
*/
if (bitOutP->bitsleft != 8) {
bitOutP->bitsleft = 1;
bit_output(bitOutP, 0); /* yes, really. This will always work. */
}
}
static void
determineBlackWhiteOrMix(const unsigned char * const image,
unsigned int const ulCol,
unsigned int const ulRow,
unsigned int const imageWidth,
unsigned int const size,
bool * const oneColorP,
int * const colorP) {
/*----------------------------------------------------------------------------
Determine whether a square within 'image' is all white, all black,
or a mix.
-----------------------------------------------------------------------------*/
unsigned int rowOfSquare;
unsigned int t;
for (rowOfSquare = 0, t = 0; rowOfSquare < size; ++rowOfSquare) {
unsigned int colOfSquare;
for (colOfSquare = 0; colOfSquare < size; ++colOfSquare) {
unsigned int rowOfImage = ulRow + rowOfSquare;
unsigned int colOfImage = ulCol + colOfSquare;
t += image[rowOfImage * imageWidth + colOfImage];
}
}
/* if the total's 0, it's black. if it's size*size, it's white. */
if (t == 0) {
*oneColorP = TRUE;
*colorP = 0;
} else if (t == SQR(size)) {
*oneColorP = TRUE;
*colorP = 1;
} else
*oneColorP = FALSE;
}
static void
doSquare(bitOut * const bitOutP,
const unsigned char * const image,
unsigned int const ulCol,
unsigned int const ulRow,
unsigned int const imageWidth,
unsigned int const size) {
/*----------------------------------------------------------------------------
Do a square of side 'size', whose upper left corner is at (ulCol, ulRow).
This is a square within 'image', which is a concatenation of rows
'imageWidth' pixels wide, one byte per pixel.
Write the pixel values out to the bit stream *bitOutP, in MRF format.
-----------------------------------------------------------------------------*/
if (size == 1) {
/* The fact that it is all one color is implied because the square is
just one pixel; no bit goes in MRF output to state that.
*/
bit_output(bitOutP, image[ulRow * imageWidth + ulCol] ? 1 : 0);
} else {
bool oneColor;
int color;
determineBlackWhiteOrMix(image, ulCol, ulRow, imageWidth, size,
&oneColor, &color);
if (oneColor) {
bit_output(bitOutP, 1); /* all same color */
bit_output(bitOutP, color);
} else {
/* Square is not all the same color, so recurse. Do each
of the four quadrants of this square individually.
*/
unsigned int const quadSize = size/2;
bit_output(bitOutP, 0); /* not all same color */
doSquare(bitOutP, image, ulCol, ulRow,
imageWidth, quadSize);
doSquare(bitOutP, image, ulCol + quadSize, ulRow,
imageWidth, quadSize);
doSquare(bitOutP, image, ulCol, ulRow + quadSize,
imageWidth, quadSize);
doSquare(bitOutP, image, ulCol + quadSize, ulRow + quadSize,
imageWidth, quadSize);
}
}
}
static void
fiddleRightEdge(unsigned char * const image,
unsigned int const w,
unsigned int const h,
unsigned int const pw,
bool * const flippedP) {
unsigned int row;
unsigned int t;
for (row = t = 0; row < h; ++row)
t += image[row*pw + w - 1];
if (t*2 > h) {
unsigned int row;
*flippedP = TRUE;
for (row = 0; row < h; ++row) {
unsigned int col;
for (col = w; col < pw; ++col)
image[row*pw + col] = 1;
}
} else
*flippedP = FALSE;
}
static void
fiddleBottomEdge(unsigned char * const image,
unsigned int const w,
unsigned int const h,
unsigned int const pw,
unsigned int const ph,
bool * const flippedP) {
unsigned int col;
unsigned int t;
for (col = t = 0; col < w; ++col)
t += image[(h-1)*pw + col];
if (t*2 > w) {
unsigned int row;
*flippedP = TRUE;
for (row = h; row < ph; ++row) {
unsigned int col;
for (col = 0; col < w; ++col)
image[row*pw + col] = 1;
}
} else
*flippedP = FALSE;
}
static void
fiddleBottomRightCorner(unsigned char * const image,
unsigned int const w,
unsigned int const h,
unsigned int const pw,
unsigned int const ph) {
unsigned int row;
for (row = h; row < ph; ++row) {
unsigned int col;
for (col = w; col < pw; ++col)
image[row*pw + col] = 1;
}
}
static void
fiddleEdges(unsigned char * const image,
int const cols,
int const rows) {
/* the aim of this routine is play around with the edges which
* are compressed into the mrf but thrown away when it's decompressed,
* such that we get the best compression possible.
* If you don't see why this is a good idea, consider the simple case
* of a 1x1 white pixel. Placed on a black 64x64 this takes several bytes
* to compress. On a white 64x64, it takes two bits.
* (Clearly most cases will be more complicated, but you should get the
* basic idea from that.)
*/
/* there are many possible approaches to this problem, and this one's
* certainly not the best, but at least it's quick and easy, and it's
* better than nothing. :-)
*
* So, all we do is flip the runoff area of an edge to white
* if more than half of the pixels on that edge are
* white. Then for the bottom-right runoff square (if there is
* one), we flip it if we flipped both edges.
*/
/* w64 is units-of-64-bits width, h64 same for height */
unsigned int const w64 = (cols + 63) / 64;
unsigned int const h64 = (rows + 63) / 64;
int const pw=w64*64;
int const ph=h64*64;
bool flippedRight, flippedBottom;
if (cols % 64 != 0)
fiddleRightEdge(image, cols, rows, pw, &flippedRight);
else
flippedRight = FALSE;
if (rows % 64 != 0)
fiddleBottomEdge(image, cols, rows, pw, ph, &flippedBottom);
else
flippedBottom = FALSE;
if (flippedRight && flippedBottom)
fiddleBottomRightCorner(image, cols, rows, pw, ph);
}
static void
readPbmImage(FILE * const ifP,
unsigned char ** const imageP,
int * const colsP,
int * const rowsP) {
/* w64 is units-of-64-bits width, h64 same for height */
unsigned int w64, h64;
unsigned char * image;
int cols, rows, format;
unsigned int row;
bit * bitrow;
pbm_readpbminit(ifP, &cols, &rows, &format);
w64 = (cols + 63) / 64;
h64 = (rows + 63) / 64;
if (UINT_MAX/w64/64/h64/64 == 0)
pm_error("Ridiculously large, unprocessable image: %u cols x %u rows",
cols, rows);
image = calloc(w64*h64*64*64, 1);
if (image == NULL)
pm_error("Unable to get memory for raster");
/* get bytemap image rounded up into mod 64x64 squares */
bitrow = pbm_allocrow(cols);
for (row = 0; row < rows; ++row) {
unsigned int col;
pbm_readpbmrow(ifP, bitrow, cols, format);
for (col =0; col < cols; ++col)
image[row*(w64*64) + col] = (bitrow[col] == PBM_WHITE ? 1 : 0);
}
pbm_freerow(bitrow);
*imageP = image;
*colsP = cols;
*rowsP = rows;
}
static void
outputMrf(FILE * const ofP,
unsigned char * const image,
unsigned int const cols,
unsigned int const rows) {
unsigned int const w64 = (cols + 63) / 64;
unsigned int const h64 = (rows + 63) / 64;
bitOut bitOut;
unsigned int row;
fprintf(ofP, "MRF1");
fprintf(ofP, "%c%c%c%c", cols >> 24, cols >> 16, cols >> 8, cols >> 0);
fprintf(ofP, "%c%c%c%c", rows >> 24, rows >> 16, rows >> 8, rows >> 0);
fputc(0, ofP); /* option byte, unused for now */
/* now recursively check squares. */
bit_init(&bitOut, ofP);
for (row = 0; row < h64; ++row) {
unsigned int col;
for (col = 0; col < w64; ++col)
doSquare(&bitOut, image, col*64, row*64, w64*64, 64);
}
bit_flush(&bitOut);
}
int
main(int argc,char *argv[]) {
FILE * ifP;
FILE * ofP;
unsigned char *image;
int rows, cols;
pbm_init(&argc, argv);
if (argc-1 > 1)
pm_error("Too many arguments: %d. Only argument is input file",
argc-1);
if (argc-1 == 1)
ifP = pm_openr(argv[1]);
else
ifP = stdin;
ofP = stdout;
readPbmImage(ifP, &image, &cols, &rows);
pm_close(ifP);
/* if necessary, alter the unused outside area to aid compression of
* edges of image.
*/
fiddleEdges(image, cols, rows);
outputMrf(ofP, image, cols, rows);
free(image);
return 0;
}
|