FFmpeg
rtjpeg.c
Go to the documentation of this file.
1 /*
2  * RTJpeg decoding functions
3  * Copyright (c) 2006 Reimar Doeffinger
4  *
5  * This file is part of FFmpeg.
6  *
7  * FFmpeg is free software; you can redistribute it and/or
8  * modify it under the terms of the GNU Lesser General Public
9  * License as published by the Free Software Foundation; either
10  * version 2.1 of the License, or (at your option) any later version.
11  *
12  * FFmpeg is distributed in the hope that it will be useful,
13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15  * Lesser General Public License for more details.
16  *
17  * You should have received a copy of the GNU Lesser General Public
18  * License along with FFmpeg; if not, write to the Free Software
19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20  */
21 #include "libavutil/common.h"
22 #include "get_bits.h"
23 #include "rtjpeg.h"
24 
25 #define PUT_COEFF(c) \
26  i = scan[coeff--]; \
27  block[i] = (c) * quant[i];
28 
29 /// aligns the bitstream to the given power of two
30 #define ALIGN(a) \
31  n = (-get_bits_count(gb)) & (a - 1); \
32  if (n) {skip_bits(gb, n);}
33 
34 /**
35  * @brief read one block from stream
36  * @param gb contains stream data
37  * @param block where data is written to
38  * @param scan array containing the mapping stream address -> block position
39  * @param quant quantization factors
40  * @return 0 means the block is not coded, < 0 means an error occurred.
41  *
42  * Note: GetBitContext is used to make the code simpler, since all data is
43  * aligned this could be done faster in a different way, e.g. as it is done
44  * in MPlayer libmpcodecs/native/rtjpegn.c.
45  */
46 static inline int get_block(GetBitContext *gb, int16_t *block, const uint8_t *scan,
47  const uint32_t *quant) {
48  int coeff, i, n;
49  int8_t ac;
50  uint8_t dc = get_bits(gb, 8);
51 
52  // block not coded
53  if (dc == 255)
54  return 0;
55 
56  // number of non-zero coefficients
57  coeff = get_bits(gb, 6);
58  if (get_bits_left(gb) < (coeff << 1))
59  return AVERROR_INVALIDDATA;
60 
61  // normally we would only need to clear the (63 - coeff) last values,
62  // but since we do not know where they are we just clear the whole block
63  memset(block, 0, 64 * sizeof(int16_t));
64 
65  // 2 bits per coefficient
66  while (coeff) {
67  ac = get_sbits(gb, 2);
68  if (ac == -2)
69  break; // continue with more bits
70  PUT_COEFF(ac);
71  }
72 
73  // 4 bits per coefficient
74  ALIGN(4);
75  if (get_bits_left(gb) < (coeff << 2))
76  return AVERROR_INVALIDDATA;
77  while (coeff) {
78  ac = get_sbits(gb, 4);
79  if (ac == -8)
80  break; // continue with more bits
81  PUT_COEFF(ac);
82  }
83 
84  // 8 bits per coefficient
85  ALIGN(8);
86  if (get_bits_left(gb) < (coeff << 3))
87  return AVERROR_INVALIDDATA;
88  while (coeff) {
89  ac = get_sbits(gb, 8);
90  PUT_COEFF(ac);
91  }
92 
93  PUT_COEFF(dc);
94  return 1;
95 }
96 
97 /**
98  * @brief decode one rtjpeg YUV420 frame
99  * @param c context, must be initialized via ff_rtjpeg_decode_init
100  * @param f AVFrame to place decoded frame into. If parts of the frame
101  * are not coded they are left unchanged, so consider initializing it
102  * @param buf buffer containing input data
103  * @param buf_size length of input data in bytes
104  * @return number of bytes consumed from the input buffer
105  */
107  const uint8_t *buf, int buf_size) {
108  GetBitContext gb;
109  int w = c->w / 16, h = c->h / 16;
110  int x, y, ret;
111  uint8_t *y1 = f->data[0], *y2 = f->data[0] + 8 * f->linesize[0];
112  uint8_t *u = f->data[1], *v = f->data[2];
113 
114  if ((ret = init_get_bits8(&gb, buf, buf_size)) < 0)
115  return ret;
116 
117  for (y = 0; y < h; y++) {
118  for (x = 0; x < w; x++) {
119 #define BLOCK(quant, dst, stride) do { \
120  int res = get_block(&gb, block, c->scan, quant); \
121  if (res < 0) \
122  return res; \
123  if (res > 0) \
124  c->idsp.idct_put(dst, stride, block); \
125 } while (0)
126  int16_t *block = c->block;
127  BLOCK(c->lquant, y1, f->linesize[0]);
128  y1 += 8;
129  BLOCK(c->lquant, y1, f->linesize[0]);
130  y1 += 8;
131  BLOCK(c->lquant, y2, f->linesize[0]);
132  y2 += 8;
133  BLOCK(c->lquant, y2, f->linesize[0]);
134  y2 += 8;
135  BLOCK(c->cquant, u, f->linesize[1]);
136  u += 8;
137  BLOCK(c->cquant, v, f->linesize[2]);
138  v += 8;
139  }
140  y1 += 2 * 8 * (f->linesize[0] - w);
141  y2 += 2 * 8 * (f->linesize[0] - w);
142  u += 8 * (f->linesize[1] - w);
143  v += 8 * (f->linesize[2] - w);
144  }
145  return get_bits_count(&gb) / 8;
146 }
147 
148 /**
149  * @brief initialize an RTJpegContext, may be called multiple times
150  * @param c context to initialize
151  * @param width width of image, will be rounded down to the nearest multiple
152  * of 16 for decoding
153  * @param height height of image, will be rounded down to the nearest multiple
154  * of 16 for decoding
155  * @param lquant luma quantization table to use
156  * @param cquant chroma quantization table to use
157  */
159  const uint32_t *lquant, const uint32_t *cquant) {
160  int i;
161  for (i = 0; i < 64; i++) {
162  int p = c->idsp.idct_permutation[i];
163  c->lquant[p] = lquant[i];
164  c->cquant[p] = cquant[i];
165  }
166  c->w = width;
167  c->h = height;
168 }
169 
171 {
172  int i;
173 
174  ff_idctdsp_init(&c->idsp, avctx);
175 
176  for (i = 0; i < 64; i++) {
177  int z = ff_zigzag_direct[i];
178  z = ((z << 3) | (z >> 3)) & 63; // rtjpeg uses a transposed variant
179 
180  // permute the scan and quantization tables for the chosen idct
181  c->scan[i] = c->idsp.idct_permutation[z];
182  }
183 }
get_bits_left
static int get_bits_left(GetBitContext *gb)
Definition: get_bits.h:849
u
#define u(width, name, range_min, range_max)
Definition: cbs_h2645.c:262
get_bits_count
static int get_bits_count(const GetBitContext *s)
Definition: get_bits.h:219
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:300
w
uint8_t w
Definition: llviddspenc.c:38
get_bits
static unsigned int get_bits(GetBitContext *s, int n)
Read 1-25 bits.
Definition: get_bits.h:379
ALIGN
#define ALIGN(a)
aligns the bitstream to the given power of two
Definition: rtjpeg.c:30
GetBitContext
Definition: get_bits.h:61
ff_rtjpeg_decode_init
void ff_rtjpeg_decode_init(RTJpegContext *c, int width, int height, const uint32_t *lquant, const uint32_t *cquant)
initialize an RTJpegContext, may be called multiple times
Definition: rtjpeg.c:158
init_get_bits8
static int init_get_bits8(GetBitContext *s, const uint8_t *buffer, int byte_size)
Initialize GetBitContext.
Definition: get_bits.h:677
width
#define width
get_sbits
static int get_sbits(GetBitContext *s, int n)
Definition: get_bits.h:359
get_bits.h
PUT_COEFF
#define PUT_COEFF(c)
Definition: rtjpeg.c:25
f
#define f(width, name)
Definition: cbs_vp9.c:255
ff_rtjpeg_decode_frame_yuv420
int ff_rtjpeg_decode_frame_yuv420(RTJpegContext *c, AVFrame *f, const uint8_t *buf, int buf_size)
decode one rtjpeg YUV420 frame
Definition: rtjpeg.c:106
c
Undefined Behavior In the C some operations are like signed integer dereferencing freed accessing outside allocated Undefined Behavior must not occur in a C it is not safe even if the output of undefined operations is unused The unsafety may seem nit picking but Optimizing compilers have in fact optimized code on the assumption that no undefined Behavior occurs Optimizing code based on wrong assumptions can and has in some cases lead to effects beyond the output of computations The signed integer overflow problem in speed critical code Code which is highly optimized and works with signed integers sometimes has the problem that often the output of the computation does not c
Definition: undefined.txt:32
dc
Tag MUST be and< 10hcoeff half pel interpolation filter coefficients, hcoeff[0] are the 2 middle coefficients[1] are the next outer ones and so on, resulting in a filter like:...eff[2], hcoeff[1], hcoeff[0], hcoeff[0], hcoeff[1], hcoeff[2] ... the sign of the coefficients is not explicitly stored but alternates after each coeff and coeff[0] is positive, so ...,+,-,+,-,+,+,-,+,-,+,... hcoeff[0] is not explicitly stored but found by subtracting the sum of all stored coefficients with signs from 32 hcoeff[0]=32 - hcoeff[1] - hcoeff[2] - ... a good choice for hcoeff and htaps is htaps=6 hcoeff={40,-10, 2} an alternative which requires more computations at both encoder and decoder side and may or may not be better is htaps=8 hcoeff={42,-14, 6,-2}ref_frames minimum of the number of available reference frames and max_ref_frames for example the first frame after a key frame always has ref_frames=1spatial_decomposition_type wavelet type 0 is a 9/7 symmetric compact integer wavelet 1 is a 5/3 symmetric compact integer wavelet others are reserved stored as delta from last, last is reset to 0 if always_reset||keyframeqlog quality(logarithmic quantizer scale) stored as delta from last, last is reset to 0 if always_reset||keyframemv_scale stored as delta from last, last is reset to 0 if always_reset||keyframe FIXME check that everything works fine if this changes between framesqbias dequantization bias stored as delta from last, last is reset to 0 if always_reset||keyframeblock_max_depth maximum depth of the block tree stored as delta from last, last is reset to 0 if always_reset||keyframequant_table quantization tableHighlevel bitstream structure:==============================--------------------------------------------|Header|--------------------------------------------|------------------------------------|||Block0||||split?||||yes no||||......... intra?||||:Block01 :yes no||||:Block02 :....... ..........||||:Block03 ::y DC ::ref index:||||:Block04 ::cb DC ::motion x :||||......... :cr DC ::motion y :||||....... ..........|||------------------------------------||------------------------------------|||Block1|||...|--------------------------------------------|------------ ------------ ------------|||Y subbands||Cb subbands||Cr subbands||||--- ---||--- ---||--- ---|||||LL0||HL0||||LL0||HL0||||LL0||HL0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||LH0||HH0||||LH0||HH0||||LH0||HH0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HL1||LH1||||HL1||LH1||||HL1||LH1|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HH1||HL2||||HH1||HL2||||HH1||HL2|||||...||...||...|||------------ ------------ ------------|--------------------------------------------Decoding process:=================------------|||Subbands|------------||||------------|Intra DC||||LL0 subband prediction ------------|\ Dequantization ------------------- \||Reference frames|\ IDWT|------- -------|Motion \|||Frame 0||Frame 1||Compensation . OBMC v -------|------- -------|--------------. \------> Frame n output Frame Frame<----------------------------------/|...|------------------- Range Coder:============Binary Range Coder:------------------- The implemented range coder is an adapted version based upon "Range encoding: an algorithm for removing redundancy from a digitised message." by G. N. N. Martin. The symbols encoded by the Snow range coder are bits(0|1). The associated probabilities are not fix but change depending on the symbol mix seen so far. bit seen|new state ---------+----------------------------------------------- 0|256 - state_transition_table[256 - old_state];1|state_transition_table[old_state];state_transition_table={ 0, 0, 0, 0, 0, 0, 0, 0, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 190, 191, 192, 194, 194, 195, 196, 197, 198, 199, 200, 201, 202, 202, 204, 205, 206, 207, 208, 209, 209, 210, 211, 212, 213, 215, 215, 216, 217, 218, 219, 220, 220, 222, 223, 224, 225, 226, 227, 227, 229, 229, 230, 231, 232, 234, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 248, 0, 0, 0, 0, 0, 0, 0};FIXME Range Coding of integers:------------------------- FIXME Neighboring Blocks:===================left and top are set to the respective blocks unless they are outside of the image in which case they are set to the Null block top-left is set to the top left block unless it is outside of the image in which case it is set to the left block if this block has no larger parent block or it is at the left side of its parent block and the top right block is not outside of the image then the top right block is used for top-right else the top-left block is used Null block y, cb, cr are 128 level, ref, mx and my are 0 Motion Vector Prediction:=========================1. the motion vectors of all the neighboring blocks are scaled to compensate for the difference of reference frames scaled_mv=(mv *(256 *(current_reference+1)/(mv.reference+1))+128)> the median of the scaled top and top right vectors is used as motion vector prediction the used motion vector is the sum of the predictor and(mvx_diff, mvy_diff) *mv_scale Intra DC Prediction block[y][x] dc[1]
Definition: snow.txt:400
ff_rtjpeg_init
void ff_rtjpeg_init(RTJpegContext *c, AVCodecContext *avctx)
Definition: rtjpeg.c:170
height
#define height
i
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:269
BLOCK
#define BLOCK(quant, dst, stride)
common.h
uint8_t
uint8_t
Definition: audio_convert.c:194
ff_idctdsp_init
av_cold void ff_idctdsp_init(IDCTDSPContext *c, AVCodecContext *avctx)
Definition: idctdsp.c:238
ff_zigzag_direct
const uint8_t ff_zigzag_direct[64]
Definition: mathtables.c:98
ret
ret
Definition: filter_design.txt:187
AVCodecContext
main external API structure.
Definition: avcodec.h:526
rtjpeg.h
quant
const uint8_t * quant
Definition: vorbis_enc_data.h:458
get_block
static int get_block(GetBitContext *gb, int16_t *block, const uint8_t *scan, const uint32_t *quant)
read one block from stream
Definition: rtjpeg.c:46
coeff
static const double coeff[2][5]
Definition: vf_owdenoise.c:72
block
The exact code depends on how similar the blocks are and how related they are to the block
Definition: filter_design.txt:207
AVERROR_INVALIDDATA
#define AVERROR_INVALIDDATA
Invalid data found when processing input.
Definition: error.h:59
h
h
Definition: vp9dsp_template.c:2038
RTJpegContext
Definition: rtjpeg.h:33