summaryrefslogtreecommitdiff
path: root/libavcodec/ljpegenc.c
blob: b6d73a42be48bc88876c3ca54581f310af56535d (plain)
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
/*
 * lossless JPEG encoder
 * Copyright (c) 2000, 2001 Fabrice Bellard
 * Copyright (c) 2003 Alex Beregszaszi
 * Copyright (c) 2003-2004 Michael Niedermayer
 *
 * Support for external huffman table, various fixes (AVID workaround),
 * aspecting, new decode_frame mechanism and apple mjpeg-b support
 *                                  by Alex Beregszaszi
 *
 * This file is part of Libav.
 *
 * Libav is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * Libav is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with Libav; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * lossless JPEG encoder.
 */

#include "libavutil/frame.h"
#include "libavutil/mem.h"
#include "libavutil/pixdesc.h"

#include "avcodec.h"
#include "idctdsp.h"
#include "internal.h"
#include "jpegtables.h"
#include "mjpegenc_common.h"
#include "mjpeg.h"
#include "mjpegenc.h"

typedef struct LJpegEncContext {
    AVClass *class;
    IDCTDSPContext idsp;
    ScanTable scantable;
    uint16_t matrix[64];

    int vsample[3];
    int hsample[3];

    uint16_t huff_code_dc_luminance[12];
    uint16_t huff_code_dc_chrominance[12];
    uint8_t  huff_size_dc_luminance[12];
    uint8_t  huff_size_dc_chrominance[12];

    uint16_t (*scratch)[4];
    int pred;
} LJpegEncContext;

static int ljpeg_encode_bgr(AVCodecContext *avctx, PutBitContext *pb,
                            const AVFrame *frame)
{
    LJpegEncContext *s    = avctx->priv_data;
    const int width       = frame->width;
    const int height      = frame->height;
    const int linesize    = frame->linesize[0];
    uint16_t (*buffer)[4] = s->scratch;
    int left[3], top[3], topleft[3];
    int x, y, i;

#if FF_API_PRIVATE_OPT
FF_DISABLE_DEPRECATION_WARNINGS
    if (avctx->prediction_method)
        s->pred = avctx->prediction_method + 1;
FF_ENABLE_DEPRECATION_WARNINGS
#endif

    for (i = 0; i < 3; i++)
        buffer[0][i] = 1 << (9 - 1);

    for (y = 0; y < height; y++) {
        const int modified_predictor = y ? s->pred : 1;
        uint8_t *ptr = frame->data[0] + (linesize * y);

        if (pb->buf_end - pb->buf - (put_bits_count(pb) >> 3) < width * 3 * 3) {
            av_log(avctx, AV_LOG_ERROR, "encoded frame too large\n");
            return -1;
        }

        for (i = 0; i < 3; i++)
            top[i]= left[i]= topleft[i]= buffer[0][i];

        for (x = 0; x < width; x++) {
            buffer[x][1] =  ptr[3 * x + 0] -     ptr[3 * x + 1] + 0x100;
            buffer[x][2] =  ptr[3 * x + 2] -     ptr[3 * x + 1] + 0x100;
            buffer[x][0] = (ptr[3 * x + 0] + 2 * ptr[3 * x + 1] + ptr[3 * x + 2]) >> 2;

            for (i = 0; i < 3; i++) {
                int pred, diff;

                PREDICT(pred, topleft[i], top[i], left[i], modified_predictor);

                topleft[i] = top[i];
                top[i]     = buffer[x+1][i];

                left[i]    = buffer[x][i];

                diff       = ((left[i] - pred + 0x100) & 0x1FF) - 0x100;

                if (i == 0)
                    ff_mjpeg_encode_dc(pb, diff, s->huff_size_dc_luminance, s->huff_code_dc_luminance); //FIXME ugly
                else
                    ff_mjpeg_encode_dc(pb, diff, s->huff_size_dc_chrominance, s->huff_code_dc_chrominance);
            }
        }
    }

    return 0;
}

static inline void ljpeg_encode_yuv_mb(LJpegEncContext *s, PutBitContext *pb,
                                       const AVFrame *frame, int predictor,
                                       int mb_x, int mb_y)
{
    int i;

    if (mb_x == 0 || mb_y == 0) {
        for (i = 0; i < 3; i++) {
            uint8_t *ptr;
            int x, y, h, v, linesize;
            h = s->hsample[i];
            v = s->vsample[i];
            linesize = frame->linesize[i];

            for (y = 0; y < v; y++) {
                for (x = 0; x < h; x++) {
                    int pred;

                    ptr = frame->data[i] + (linesize * (v * mb_y + y)) + (h * mb_x + x); //FIXME optimize this crap
                    if (y == 0 && mb_y == 0) {
                        if (x == 0 && mb_x == 0)
                            pred = 128;
                        else
                            pred = ptr[-1];
                    } else {
                        if (x == 0 && mb_x == 0) {
                            pred = ptr[-linesize];
                        } else {
                            PREDICT(pred, ptr[-linesize - 1], ptr[-linesize],
                                    ptr[-1], predictor);
                        }
                    }

                    if (i == 0)
                        ff_mjpeg_encode_dc(pb, *ptr - pred, s->huff_size_dc_luminance, s->huff_code_dc_luminance); //FIXME ugly
                    else
                        ff_mjpeg_encode_dc(pb, *ptr - pred, s->huff_size_dc_chrominance, s->huff_code_dc_chrominance);
                }
            }
        }
    } else {
        for (i = 0; i < 3; i++) {
            uint8_t *ptr;
            int x, y, h, v, linesize;
            h = s->hsample[i];
            v = s->vsample[i];
            linesize = frame->linesize[i];

            for (y = 0; y < v; y++) {
                for (x = 0; x < h; x++) {
                    int pred;

                    ptr = frame->data[i] + (linesize * (v * mb_y + y)) + (h * mb_x + x); //FIXME optimize this crap
                    PREDICT(pred, ptr[-linesize - 1], ptr[-linesize], ptr[-1], predictor);

                    if (i == 0)
                        ff_mjpeg_encode_dc(pb, *ptr - pred, s->huff_size_dc_luminance, s->huff_code_dc_luminance); //FIXME ugly
                    else
                        ff_mjpeg_encode_dc(pb, *ptr - pred, s->huff_size_dc_chrominance, s->huff_code_dc_chrominance);
                }
            }
        }
    }
}

static int ljpeg_encode_yuv(AVCodecContext *avctx, PutBitContext *pb,
                            const AVFrame *frame)
{
    LJpegEncContext *s  = avctx->priv_data;
    const int mb_width  = (avctx->width  + s->hsample[0] - 1) / s->hsample[0];
    const int mb_height = (avctx->height + s->vsample[0] - 1) / s->vsample[0];
    int mb_x, mb_y;

#if FF_API_PRIVATE_OPT
FF_DISABLE_DEPRECATION_WARNINGS
    if (avctx->prediction_method)
        s->pred = avctx->prediction_method + 1;
FF_ENABLE_DEPRECATION_WARNINGS
#endif

    for (mb_y = 0; mb_y < mb_height; mb_y++) {
        if (pb->buf_end - pb->buf - (put_bits_count(pb) >> 3) <
            mb_width * 4 * 3 * s->hsample[0] * s->vsample[0]) {
            av_log(avctx, AV_LOG_ERROR, "encoded frame too large\n");
            return -1;
        }

        for (mb_x = 0; mb_x < mb_width; mb_x++)
            ljpeg_encode_yuv_mb(s, pb, frame, s->pred, mb_x, mb_y);
    }

    return 0;
}

static int ljpeg_encode_frame(AVCodecContext *avctx, AVPacket *pkt,
                              const AVFrame *pict, int *got_packet)
{
    LJpegEncContext *s = avctx->priv_data;
    PutBitContext pb;
    const int width  = avctx->width;
    const int height = avctx->height;
    const int mb_width  = (width  + s->hsample[0] - 1) / s->hsample[0];
    const int mb_height = (height + s->vsample[0] - 1) / s->vsample[0];
    int max_pkt_size = AV_INPUT_BUFFER_MIN_SIZE;
    int ret, header_bits;

    if (avctx->pix_fmt == AV_PIX_FMT_BGR24)
        max_pkt_size += width * height * 3 * 3;
    else {
        max_pkt_size += mb_width * mb_height * 3 * 4
                        * s->hsample[0] * s->vsample[0];
    }
    if ((ret = ff_alloc_packet(pkt, max_pkt_size)) < 0) {
        av_log(avctx, AV_LOG_ERROR, "Error getting output packet of size %d.\n", max_pkt_size);
        return ret;
    }

    init_put_bits(&pb, pkt->data, pkt->size);

    ff_mjpeg_encode_picture_header(avctx, &pb, &s->scantable,
                                   s->pred, s->matrix);

    header_bits = put_bits_count(&pb);

    if (avctx->pix_fmt == AV_PIX_FMT_BGR24)
        ret = ljpeg_encode_bgr(avctx, &pb, pict);
    else
        ret = ljpeg_encode_yuv(avctx, &pb, pict);
    if (ret < 0)
        return ret;

    emms_c();

    ff_mjpeg_encode_picture_trailer(&pb, header_bits);

    flush_put_bits(&pb);
    pkt->size   = put_bits_ptr(&pb) - pb.buf;
    pkt->flags |= AV_PKT_FLAG_KEY;
    *got_packet = 1;

    return 0;
}

static av_cold int ljpeg_encode_close(AVCodecContext *avctx)
{
    LJpegEncContext *s = avctx->priv_data;

    av_freep(&s->scratch);

    return 0;
}

static av_cold int ljpeg_encode_init(AVCodecContext *avctx)
{
    LJpegEncContext *s = avctx->priv_data;
    int chroma_v_shift, chroma_h_shift;

    if ((avctx->pix_fmt == AV_PIX_FMT_YUV420P ||
         avctx->pix_fmt == AV_PIX_FMT_YUV422P ||
         avctx->pix_fmt == AV_PIX_FMT_YUV444P ||
         avctx->color_range == AVCOL_RANGE_MPEG) &&
        avctx->strict_std_compliance > FF_COMPLIANCE_UNOFFICIAL) {
        av_log(avctx, AV_LOG_ERROR,
               "Limited range YUV is non-standard, set strict_std_compliance to "
               "at least unofficial to use it.\n");
        return AVERROR(EINVAL);
    }

#if FF_API_CODED_FRAME
FF_DISABLE_DEPRECATION_WARNINGS
    avctx->coded_frame->pict_type = AV_PICTURE_TYPE_I;
    avctx->coded_frame->key_frame = 1;
FF_ENABLE_DEPRECATION_WARNINGS
#endif

    s->scratch = av_malloc_array(avctx->width + 1, sizeof(*s->scratch));

    ff_idctdsp_init(&s->idsp, avctx);
    ff_init_scantable(s->idsp.idct_permutation, &s->scantable,
                      ff_zigzag_direct);

    av_pix_fmt_get_chroma_sub_sample(avctx->pix_fmt, &chroma_h_shift,
                                     &chroma_v_shift);

    if (avctx->pix_fmt   == AV_PIX_FMT_BGR24) {
        s->vsample[0] = s->hsample[0] =
        s->vsample[1] = s->hsample[1] =
        s->vsample[2] = s->hsample[2] = 1;
    } else {
        s->vsample[0] = 2;
        s->vsample[1] = 2 >> chroma_v_shift;
        s->vsample[2] = 2 >> chroma_v_shift;
        s->hsample[0] = 2;
        s->hsample[1] = 2 >> chroma_h_shift;
        s->hsample[2] = 2 >> chroma_h_shift;
    }

    ff_mjpeg_build_huffman_codes(s->huff_size_dc_luminance,
                                 s->huff_code_dc_luminance,
                                 avpriv_mjpeg_bits_dc_luminance,
                                 avpriv_mjpeg_val_dc);
    ff_mjpeg_build_huffman_codes(s->huff_size_dc_chrominance,
                                 s->huff_code_dc_chrominance,
                                 avpriv_mjpeg_bits_dc_chrominance,
                                 avpriv_mjpeg_val_dc);

    return 0;
}

#define OFFSET(x) offsetof(LJpegEncContext, x)
#define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
static const AVOption options[] = {
{ "pred", "Prediction method", OFFSET(pred), AV_OPT_TYPE_INT, { .i64 = 1 }, 1, 3, VE, "pred" },
    { "left",   NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 1 }, INT_MIN, INT_MAX, VE, "pred" },
    { "plane",  NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 2 }, INT_MIN, INT_MAX, VE, "pred" },
    { "median", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 3 }, INT_MIN, INT_MAX, VE, "pred" },

    { NULL},
};

static const AVClass ljpeg_class = {
    .class_name = "ljpeg",
    .item_name  = av_default_item_name,
    .option     = options,
    .version    = LIBAVUTIL_VERSION_INT,
};

AVCodec ff_ljpeg_encoder = {
    .name           = "ljpeg",
    .long_name      = NULL_IF_CONFIG_SMALL("Lossless JPEG"),
    .type           = AVMEDIA_TYPE_VIDEO,
    .id             = AV_CODEC_ID_LJPEG,
    .priv_data_size = sizeof(LJpegEncContext),
    .priv_class     = &ljpeg_class,
    .init           = ljpeg_encode_init,
    .encode2        = ljpeg_encode_frame,
    .close          = ljpeg_encode_close,
    .pix_fmts       = (const enum AVPixelFormat[]){ AV_PIX_FMT_YUVJ420P,
                                                    AV_PIX_FMT_YUVJ422P,
                                                    AV_PIX_FMT_YUVJ444P,
                                                    AV_PIX_FMT_BGR24,
                                                    AV_PIX_FMT_YUV420P,
                                                    AV_PIX_FMT_YUV422P,
                                                    AV_PIX_FMT_YUV444P,
                                                    AV_PIX_FMT_NONE },
};