发新帖我要提问
返回列表
打印

纯C实现的JPEG压缩算法

[复制链接]
1985|8
手机看帖
扫描二维码
随时随地手机跟帖
跳转到指定楼层
楼主
small泡泡|  楼主 | 2014-1-10 17:26 | 只看该作者 回帖奖励 |倒序浏览 |阅读模式
JPEG压缩, ui, TPU, Output, Input
  
#include <stdio.h>
typedef unsigned int UINT32;
typedef int    INT32;
typedef unsigned short UINT16;
typedef short   INT16;
typedef unsigned char UINT8;
typedef char   INT8;
#define CLIP(color) (unsigned char)(((color)>0xFF)?0xff:(((color)<0)?0:(color)))
#define  BLOCK_SIZE 64  
typedef struct JPEG_ENCODER_STRUCTURE
{
UINT16 mcu_width;
UINT16 mcu_height;
UINT16 horizontal_mcus;
UINT16 vertical_mcus;
UINT16 rows_in_bottom_mcus;
UINT16 cols_in_right_mcus;
UINT16 length_minus_mcu_width;
UINT16 length_minus_width;
UINT16 mcu_width_size;
UINT16 offset;
INT16 ldc1;
INT16 ldc2;
INT16 ldc3;
UINT16 rows;
UINT16 cols;
UINT16 incr;
} JPEG_ENCODER_STRUCTURE;
UINT8  Lqt[BLOCK_SIZE];
UINT8  Cqt[BLOCK_SIZE];
UINT16  ILqt[BLOCK_SIZE];
UINT16  ICqt[BLOCK_SIZE];
INT16  Y1[BLOCK_SIZE];
INT16  Y2[BLOCK_SIZE];
INT16  Y3[BLOCK_SIZE];
INT16  Y4[BLOCK_SIZE];
INT16  CB[BLOCK_SIZE];
INT16  CR[BLOCK_SIZE];
INT16  Temp[BLOCK_SIZE];
INT32  lcode;
UINT16  bitindex;
static UINT8 zigzag_table[] = {};
// Header for JPEG Encoder
static UINT16 markerdata[] = {
    // dht
    0xFFC4, 0x1A2, 0x00,
    // luminance dc (2 - 16) + 1
    0x0105, 0x0101, 0x00101, 0x0101, 0x0000, 0x00000, 00000, 00000,
    // luminance dc (2 - 12) + 1
    0x0102, 0x0304, 0x0506, 0x0708, 0x090A, 0x0B01,
    // chrominance dc (1 - 16)
    0x0003, 0x0101, 0x0101, 0x0101, 0x0101, 0x0100, 0x0000, 0x0000,
    // chrominance dc (1 - 12)
    0x0001, 0x00203, 0x0405, 0x0607, 0x0809, 0x00A0B,
    // luminance ac 1 + (1 - 15)
    0x1000, 0x0201, 0x0303, 0x0204, 0x0305, 0x0504, 0x0400, 0x0001,
    // luminance ac 1 + (1 - 162) + 1

    // chrominance ac (1 - 16)
    0x0002, 0x0102, 0x0404, 0x0304, 0x0705, 0x0404, 0x0001, 0x0277,
    // chrominance ac (1 - 162)};
UINT8 bitsize[] = {};
UINT16 luminance_dc_code_table[] = {
0x0000, 0x0002, 0x0003, 0x0004, 0x0005, 0x0006, 0x000E, 0x001E, 0x003E,
    0x007E, 0x00FE, 0x01FE
};
UINT16 luminance_dc_size_table[] = {
0x0002, 0x0003, 0x0003, 0x0003, 0x0003, 0x0003, 0x0004, 0x0005, 0x0006,
0x0007, 0x0008, 0x0009
};
UINT16 chrominance_dc_code_table[] = {
0x0000, 0x0001, 0x0002, 0x0006, 0x000E, 0x001E, 0x003E, 0x007E, 0x00FE,
0x01FE, 0x03FE, 0x07FE
};
UINT16 chrominance_dc_size_table[] = {
0x0002, 0x0002, 0x0002, 0x0003, 0x0004, 0x0005, 0x0006, 0x0007, 0x0008,
0x0009, 0x000A, 0x000B
};
UINT16 luminance_ac_code_table[] = {};
UINT16 luminance_ac_size_table[] = {};
UINT16 chrominance_ac_code_table[] = {};
UINT16 chrominance_ac_size_table[] = {};
#define  FOUR_ZERO_ZERO 0 // Grey scale Y00 ...
#define  FOUR_TWO_TWO 2 //Y00 Cb Y01 Cr
static void (*read_format) (JPEG_ENCODER_STRUCTURE * jpeg_encoder_structure,UINT8 * input_ptr);
static void RGB_2_400 (UINT8 * input_ptr, UINT8 * output_ptr, UINT32 image_width,UINT32 image_height)
{
  UINT32 i, size;
  UINT8 R, G, B;
  INT32 Y;
  UINT8 * inbuf = input_ptr;
  size = image_width * image_height;
  for (i = size; i > 0; i--)
   
    {
      B = inbuf[0];
      G = inbuf[1];
      R = inbuf[2];
      inbuf += 3;
      Y = CLIP ((77 * R + 150 * G + 29 * B) >> 8);
      *input_ptr++ = (UINT8) Y;
    }
}
static void read_400_format (JPEG_ENCODER_STRUCTURE * jpeg_encoder_structure,UINT8 * input_ptr)
{
INT32 i, j;
INT16 * Y1_Ptr = Y1;
UINT16 rows = jpeg_encoder_structure->rows;
UINT16 cols = jpeg_encoder_structure->cols;
UINT16 incr = jpeg_encoder_structure->incr;
for (i = rows; i > 0; i--)
{
  for (j = cols; j > 0; j--)
   *Y1_Ptr++ = *input_ptr++ - 128;
  for (j = 8 - cols; j > 0; j--)
   *Y1_Ptr++ = *(Y1_Ptr - 1);
  input_ptr += incr;
}
for (i = 8 - rows; i > 0; i--)
{
  for (j = 8; j > 0; j--)
   *Y1_Ptr++ = *(Y1_Ptr - 8);
}
}
static void
RGB_2_422 (UINT8 * input_ptr, UINT8 * output_ptr, UINT32 image_width,
     UINT32 image_height)
{
UINT32 i, size;
UINT8 R, G, B, R1, G1, B1;
INT32 Y, Yp, Cb, Cr;
UINT8 * inbuf = input_ptr;
size = image_width * image_height/2;
for (i = size; i > 0; i--)
    {
  B = inbuf[0];
  G = inbuf[1];
  R = inbuf[2];
  B1 = inbuf[3];
  G1 = inbuf[4];
  R1 = inbuf[5];
  inbuf += 6;
  Y = CLIP ((77 * R + 150 * G + 29 * B) >> 8);
  Yp = CLIP ((77 * R1 + 150 * G1 + 29 * B1) >> 8);
  Cb = CLIP (((-43 * R - 85 * G + 128 * B) >> 8) + 128);
  Cr = CLIP (((128 * R - 107 * G - 21 * B) >> 8) + 128);
  *input_ptr++ = (UINT8) Y;
  *input_ptr++ = (UINT8) Cb;
  *input_ptr++ = (UINT8) Yp;
  *input_ptr++ = (UINT8) Cr;
    }
}
/* This function implements 16 Step division for Q.15 format data */
UINT16 DSP_Division (UINT32 numer, UINT32 denom)
{
UINT16 i;
denom <<= 15;
for (i = 16; i > 0; i--)
{
  if (numer > denom)
  {
   numer -= denom;
   numer <<= 1;
   numer++;
  }
  else
   numer <<= 1;
}
return (UINT16) numer;
}
UINT8 * write_markers (UINT8 * output_ptr, UINT32 image_format,UINT32 image_width, UINT32 image_height)
{
UINT16 i, header_length;
UINT8 number_of_components;
// Start of image marker
*output_ptr++ = 0xFF;
*output_ptr++ = 0xD8;
// Quantization table marker
*output_ptr++ = 0xFF;
*output_ptr++ = 0xDB;
// Quantization table length
*output_ptr++ = 0x00;
*output_ptr++ = 0x84;
// Pq, Tq
*output_ptr++ = 0x00;
// Lqt table
for (i = 0; i < 64; i++)
  *output_ptr++ = Lqt[i];
// Pq, Tq
*output_ptr++ = 0x01;
// Cqt table
for (i = 0; i < 64; i++)
  *output_ptr++ = Cqt[i];
// huffman table(DHT)
for (i = 0; i < 210; i++)
{
  *output_ptr++ = (UINT8) (markerdata[i] >> 8);
  *output_ptr++ = (UINT8) markerdata[i];
}
//FOUR_ZERO_ZERO
number_of_components = 3;
// Frame header(SOF)
// Start of frame marker
*output_ptr++ = 0xFF;
*output_ptr++ = 0xC0;
header_length = (UINT16) (8 + 3 * number_of_components);
// Frame header length  
*output_ptr++ = (UINT8) (header_length >> 8);
*output_ptr++ = (UINT8) header_length;
// Precision (P)
*output_ptr++ = 0x08;
// image height
*output_ptr++ = (UINT8) (image_height >> 8);
*output_ptr++ = (UINT8) image_height;
// image width
*output_ptr++ = (UINT8) (image_width >> 8);
*output_ptr++ = (UINT8) image_width;
// Nf
*output_ptr++ = number_of_components;
// FOUR_ZERO_ZERO
      *output_ptr++ = 0x01;
    *output_ptr++ = 0x21;
      *output_ptr++ = 0x00;
      *output_ptr++ = 0x02;
      *output_ptr++ = 0x11;
      *output_ptr++ = 0x01;
      *output_ptr++ = 0x03;
      *output_ptr++ = 0x11;
      *output_ptr++ = 0x01;

// Scan header(SOF)
// Start of scan marker
*output_ptr++ = 0xFF;
*output_ptr++ = 0xDA;
header_length = (UINT16) (6 + (number_of_components << 1));
// Scan header length
*output_ptr++ = (UINT8) (header_length >> 8);
*output_ptr++ = (UINT8) header_length;
// Ns
*output_ptr++ = number_of_components;
//FOUR_ZERO_ZERO
      *output_ptr++ = 0x01;
      *output_ptr++ = 0x00;
      *output_ptr++ = 0x02;
      *output_ptr++ = 0x11;
      *output_ptr++ = 0x03;
      *output_ptr++ = 0x11;
*output_ptr++ = 0x00;
*output_ptr++ = 0x3F;
*output_ptr++ = 0x00;
return output_ptr;
}
static void
read_422_format (JPEG_ENCODER_STRUCTURE * jpeg_encoder_structure,
     UINT8 * input_ptr)
{
INT32 i, j;
UINT16 Y1_cols, Y2_cols;
INT16 * Y1_Ptr = Y1;
INT16 * Y2_Ptr = Y2;
INT16 * CB_Ptr = CB;
INT16 * CR_Ptr = CR;
UINT16 rows = jpeg_encoder_structure->rows;
UINT16 cols = jpeg_encoder_structure->cols;
UINT16 incr = jpeg_encoder_structure->incr;
if (cols <= 8)
  
    {
  Y1_cols = cols;
  Y2_cols = 0;
    }

else
  
    {
  Y1_cols = 8;
  Y2_cols = (UINT16) (cols - 8);
    }
for (i = rows; i > 0; i--)
  
    {
  for (j = Y1_cols >> 1; j > 0; j--)
   
  {
   *Y1_Ptr++ = *input_ptr++ - 128;
   *CB_Ptr++ = *input_ptr++ - 128;
   *Y1_Ptr++ = *input_ptr++ - 128;
   *CR_Ptr++ = *input_ptr++ - 128;
  }
  for (j = Y2_cols >> 1; j > 0; j--)
   
  {
   *Y2_Ptr++ = *input_ptr++ - 128;
   *CB_Ptr++ = *input_ptr++ - 128;
   *Y2_Ptr++ = *input_ptr++ - 128;
   *CR_Ptr++ = *input_ptr++ - 128;
  }
  if (cols <= 8)
   
  {
   for (j = 8 - Y1_cols; j > 0; j--)
    *Y1_Ptr++ = *(Y1_Ptr - 1);
   for (j = 8 - Y2_cols; j > 0; j--)
    *Y2_Ptr++ = *(Y1_Ptr - 1);
  }
  
  else
   
  {
   for (j = 8 - Y2_cols; j > 0; j--)
    *Y2_Ptr++ = *(Y2_Ptr - 1);
  }
  for (j = (16 - cols) >> 1; j > 0; j--)
   
  {
   *CB_Ptr++ = *(CB_Ptr - 1);
   *CR_Ptr++ = *(CR_Ptr - 1);
  }
  input_ptr += incr;
    }
for (i = 8 - rows; i > 0; i--)
  
    {
  for (j = 8; j > 0; j--)
   
  {
   *Y1_Ptr++ = *(Y1_Ptr - 8);
   *Y2_Ptr++ = *(Y2_Ptr - 8);
   *CB_Ptr++ = *(CB_Ptr - 8);
   *CR_Ptr++ = *(CR_Ptr - 8);
  }
    }
}

使用特权

评论回复 赏
评分
参与人数 1威望 +2 收起 理由
地瓜patch + 2 神马都是浮云

相关下载

相关帖子
沙发
small泡泡|  楼主 | 2014-1-10 17:27 | 只看该作者
static void initialization (JPEG_ENCODER_STRUCTURE * jpeg, UINT32 image_format,UINT32 image_width, UINT32 image_height)
{
UINT16 mcu_width, mcu_height, bytes_per_pixel;
lcode = 0;
bitindex = 0;
      jpeg->mcu_width = mcu_width = 16;
      jpeg->horizontal_mcus = (UINT16) ((image_width + mcu_width - 1) >> 4);
   jpeg->mcu_height = mcu_height = 8;
   jpeg->vertical_mcus =
    (UINT16) ((image_height + mcu_height - 1) >> 3);
   bytes_per_pixel = 2;
   read_format = read_422_format;
   
jpeg->rows_in_bottom_mcus =(UINT16) (image_height - (jpeg->vertical_mcus - 1) * mcu_height);
jpeg->cols_in_right_mcus =(UINT16) (image_width - (jpeg->horizontal_mcus - 1) * mcu_width);
jpeg->length_minus_mcu_width =(UINT16) ((image_width - mcu_width) * bytes_per_pixel);
jpeg->length_minus_width =(UINT16) ((image_width - jpeg->cols_in_right_mcus) * bytes_per_pixel);
jpeg->mcu_width_size = (UINT16) (mcu_width * bytes_per_pixel);
jpeg->offset =(UINT16) ((image_width * (mcu_height - 1) -(mcu_width - jpeg->cols_in_right_mcus)) * bytes_per_pixel);
jpeg->ldc1 = 0;
jpeg->ldc2 = 0;
jpeg->ldc3 = 0;
}
/* Multiply Quantization table with quality factor to get LQT and CQT */
void initialize_quantization_tables (UINT32 quality_factor)
{
UINT16 i, index;
UINT32 value;
UINT8 luminance_quant_table[] = {
16, 11, 10, 16, 24, 40, 51, 61, 12, 12, 14, 19, 26, 58, 60, 55, 14, 13,
16, 24, 40, 57, 69, 56, 14, 17, 22, 29, 51, 87, 80, 62, 18, 22, 37,
56, 68, 109, 103, 77, 24, 35, 55, 64, 81, 104, 113, 92, 49, 64, 78,
87, 103, 121, 120, 101, 72, 92, 95, 98, 112, 100, 103, 99
};
UINT8 chrominance_quant_table[] = {
17, 18, 24, 47, 99, 99, 99, 99, 18, 21, 26, 66, 99, 99, 99, 99, 24, 26,
56, 99, 99, 99, 99, 99, 47, 66, 99, 99, 99, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99
};
for (i = 0; i < 64; i++)
{
  index = zigzag_table[i];
  /* luminance quantization table * quality factor */
  value = luminance_quant_table[i] * quality_factor;
  value = (value + 0x200) >> 10;
  if (value == 0)
  value = 1;
  else if (value > 255)
  value = 255;
  Lqt[index] = (UINT8) value;
  ILqt[i] = DSP_Division (0x8000, value);
  /* chrominance quantization table * quality factor */
  value = chrominance_quant_table[i] * quality_factor;
  value = (value + 0x200) >> 10;
  if (value == 0)
  value = 1;
  else if (value > 255)
  value = 255;
  Cqt[index] = (UINT8) value;
  ICqt[i] = DSP_Division (0x8000, value);
}
}
UINT8 * close_bitstream (UINT8 * output_ptr)
{
UINT16 i, count;
UINT8 * ptr;
if (bitindex > 0)
{
  lcode <<= (32 - bitindex);
  count = (bitindex + 7) >> 3;
  ptr = (UINT8 *) & lcode + 3;
  for (i = count; i > 0; i--)
  {
   if ((*output_ptr++ = *ptr--) == 0xff)
   *output_ptr++ = 0;
  }
}
// End of image marker
*output_ptr++ = 0xFF;
*output_ptr++ = 0xD9;
return output_ptr;
}
/* DCT for One block(8x8) */
static void DCT (INT16 * data)
{
  UINT16 i;
  INT32 x0, x1, x2, x3, x4, x5, x6, x7, x8;
  
/* All values are shifted left by 10
and rounded off to nearest integer */
  static const UINT16 c1 = 1420; /* cos PI/16 * root(2)  */
  static const UINT16 c2 = 1338; /* cos PI/8 * root(2)   */
  static const UINT16 c3 = 1204; /* cos 3PI/16 * root(2) */
  static const UINT16 c5 = 805; /* cos 5PI/16 * root(2) */
  static const UINT16 c6 = 554; /* cos 3PI/8 * root(2)  */
  static const UINT16 c7 = 283; /* cos 7PI/16 * root(2) */
  static const UINT16 s1 = 3;
  static const UINT16 s2 = 10;
  static const UINT16 s3 = 13;
  for (i = 8; i > 0; i--)
   
    {
      x8 = data[0] + data[7];
      x0 = data[0] - data[7];
      x7 = data[1] + data[6];
      x1 = data[1] - data[6];
      x6 = data[2] + data[5];
      x2 = data[2] - data[5];
      x5 = data[3] + data[4];
      x3 = data[3] - data[4];
      x4 = x8 + x5;
      x8 -= x5;
      x5 = x7 + x6;
      x7 -= x6;
      data[0] = (INT16) (x4 + x5);
      data[4] = (INT16) (x4 - x5);
      data[2] = (INT16) ((x8 * c2 + x7 * c6) >> s2);
      data[6] = (INT16) ((x8 * c6 - x7 * c2) >> s2);
      data[7] = (INT16) ((x0 * c7 - x1 * c5 + x2 * c3 - x3 * c1) >> s2);
      data[5] = (INT16) ((x0 * c5 - x1 * c1 + x2 * c7 + x3 * c3) >> s2);
      data[3] = (INT16) ((x0 * c3 - x1 * c7 - x2 * c1 - x3 * c5) >> s2);
      data[1] = (INT16) ((x0 * c1 + x1 * c3 + x2 * c5 + x3 * c7) >> s2);
      data += 8;
    }
  data -= 64;
  for (i = 8; i > 0; i--)
   
    {
      x8 = data[0] + data[56];
      x0 = data[0] - data[56];
      x7 = data[8] + data[48];
      x1 = data[8] - data[48];
      x6 = data[16] + data[40];
      x2 = data[16] - data[40];
      x5 = data[24] + data[32];
      x3 = data[24] - data[32];
      x4 = x8 + x5;
      x8 -= x5;
      x5 = x7 + x6;
      x7 -= x6;
      data[0] = (INT16) ((x4 + x5) >> s1);
      data[32] = (INT16) ((x4 - x5) >> s1);
      data[16] = (INT16) ((x8 * c2 + x7 * c6) >> s3);
      data[48] = (INT16) ((x8 * c6 - x7 * c2) >> s3);
      data[56] = (INT16) ((x0 * c7 - x1 * c5 + x2 * c3 - x3 * c1) >> s3);
      data[40] = (INT16) ((x0 * c5 - x1 * c1 + x2 * c7 + x3 * c3) >> s3);
      data[24] = (INT16) ((x0 * c3 - x1 * c7 - x2 * c1 - x3 * c5) >> s3);
      data[8] = (INT16) ((x0 * c1 + x1 * c3 + x2 * c5 + x3 * c7) >> s3);
      data++;
    }
}
/* multiply DCT Coefficients with Quantization table and store in ZigZag location */
void quantization (INT16 * const data, UINT16 * const quant_table_ptr)
{
  INT16 i;
  INT32 value;
  for (i = 63; i >= 0; i--)
   
    {
      value = data[i] * quant_table_ptr[i];
      value = (value + 0x4000) >> 15;
      Temp[zigzag_table[i]] = (INT16) value;
    }
}

//huffman算法
UINT8 * huffman (JPEG_ENCODER_STRUCTURE * jpeg_encoder_structure,UINT16 component, UINT8 * output_ptr)
{
  UINT16 i;
  UINT16 * DcCodeTable, *DcSizeTable, *AcCodeTable, *AcSizeTable;
  INT16 * Temp_Ptr, Coeff, LastDc;
  UINT16 AbsCoeff, HuffCode, HuffSize, RunLength = 0, DataSize = 0, index;
  INT16 bits_in_next_word;
  UINT16 numbits;
  UINT32 data;
  Temp_Ptr = Temp;
  Coeff = *Temp_Ptr++;
  if (component == 1)
   
    {
      DcCodeTable = luminance_dc_code_table;
      DcSizeTable = luminance_dc_size_table;
      AcCodeTable = luminance_ac_code_table;
      AcSizeTable = luminance_ac_size_table;
      LastDc = jpeg_encoder_structure->ldc1;
      jpeg_encoder_structure->ldc1 = Coeff;
    }
  
  else
   
    {
      DcCodeTable = chrominance_dc_code_table;
      DcSizeTable = chrominance_dc_size_table;
      AcCodeTable = chrominance_ac_code_table;
      AcSizeTable = chrominance_ac_size_table;
      if (component == 2)

{
   LastDc = jpeg_encoder_structure->ldc2;
   jpeg_encoder_structure->ldc2 = Coeff;
}
      
      else

{
   LastDc = jpeg_encoder_structure->ldc3;
   jpeg_encoder_structure->ldc3 = Coeff;
}
    }
  Coeff -= LastDc;
  AbsCoeff = (Coeff < 0) ? -Coeff-- : Coeff;
  while (AbsCoeff != 0)
   
    {
      AbsCoeff >>= 1;
      DataSize++;
    }
  HuffCode = DcCodeTable[DataSize];
  HuffSize = DcSizeTable[DataSize];
  Coeff &= (1 << DataSize) - 1;
  data = (HuffCode << DataSize) | Coeff;
  numbits = HuffSize + DataSize;
  //PUTBITS
  {
  bits_in_next_word = (INT16) (bitindex + numbits - 32);
  if (bits_in_next_word < 0)
    {
    lcode = (lcode << numbits) | data;
    bitindex += numbits;
    }
  else
    {
    lcode = (lcode << (32 - bitindex)) | (data >> bits_in_next_word);
    if ((*output_ptr++ = (UINT8) (lcode >> 24)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 16)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 8)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) lcode) == 0xff)
    *output_ptr++ = 0;
    lcode = data;
    bitindex = bits_in_next_word;
    }
    }  
  for (i = 63; i > 0; i--)
   
    {
      if ((Coeff = *Temp_Ptr++) != 0)

{
   while (RunLength > 15)
     
     {
       RunLength -= 16;
       data = AcCodeTable[161];
       numbits = AcSizeTable[161];
     //PUTBITS
     {
  bits_in_next_word = (INT16) (bitindex + numbits - 32);
  if (bits_in_next_word < 0)
    {
    lcode = (lcode << numbits) | data;
    bitindex += numbits;
    }
  else
    {
    lcode = (lcode << (32 - bitindex)) | (data >> bits_in_next_word);
    if ((*output_ptr++ = (UINT8) (lcode >> 24)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 16)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 8)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) lcode) == 0xff)
    *output_ptr++ = 0;
    lcode = data;
    bitindex = bits_in_next_word;
    }
    }   
      }
   AbsCoeff = (Coeff < 0) ? -Coeff-- : Coeff;
   if (AbsCoeff >> 8 == 0)
     DataSize = bitsize[AbsCoeff];
   
   else
   DataSize = bitsize[AbsCoeff >> 8] + 8;
   index = RunLength * 10 + DataSize;
   HuffCode = AcCodeTable[index];
   HuffSize = AcSizeTable[index];
   Coeff &= (1 << DataSize) - 1;
   data = (HuffCode << DataSize) | Coeff;
   numbits = HuffSize + DataSize;
  // PUTBITS
    {
  bits_in_next_word = (INT16) (bitindex + numbits - 32);
  if (bits_in_next_word < 0)
    {
    lcode = (lcode << numbits) | data;
    bitindex += numbits;
    }
  else
    {
    lcode = (lcode << (32 - bitindex)) | (data >> bits_in_next_word);
    if ((*output_ptr++ = (UINT8) (lcode >> 24)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 16)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 8)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) lcode) == 0xff)
    *output_ptr++ = 0;
    lcode = data;
    bitindex = bits_in_next_word;
    }
    }   
   RunLength = 0;
}
      
      else
RunLength++;
    }
  if (RunLength != 0)
   
    {
      data = AcCodeTable[0];
      numbits = AcSizeTable[0];
    // PUTBITS
     {
  bits_in_next_word = (INT16) (bitindex + numbits - 32);
  if (bits_in_next_word < 0)
    {
    lcode = (lcode << numbits) | data;
    bitindex += numbits;
    }
  else
    {
    lcode = (lcode << (32 - bitindex)) | (data >> bits_in_next_word);
    if ((*output_ptr++ = (UINT8) (lcode >> 24)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 16)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) (lcode >> 8)) == 0xff)
    *output_ptr++ = 0;
    if ((*output_ptr++ = (UINT8) lcode) == 0xff)
    *output_ptr++ = 0;
    lcode = data;
    bitindex = bits_in_next_word;
    }
    }  
    }
  return output_ptr;
}

static UINT8 *encodeMCU (JPEG_ENCODER_STRUCTURE * jpeg_encoder_structure,UINT32 image_format, UINT8 * output_ptr)
{
  DCT (Y1);
  quantization (Y1, ILqt);
  output_ptr = huffman (jpeg_encoder_structure, 1, output_ptr);
  if (image_format == FOUR_ZERO_ZERO)
   return output_ptr;
  DCT (Y2);
  quantization (Y2, ILqt);
  output_ptr = huffman (jpeg_encoder_structure, 1, output_ptr);
  if (image_format == FOUR_TWO_TWO)
   goto chroma;
  DCT (Y3);
  quantization (Y3, ILqt);
  output_ptr = huffman (jpeg_encoder_structure, 1, output_ptr);
  DCT (Y4);
  quantization (Y4, ILqt);
  output_ptr = huffman (jpeg_encoder_structure, 1, output_ptr);
chroma:DCT (CB);
    quantization (CB, ICqt);
    output_ptr = huffman (jpeg_encoder_structure, 2, output_ptr);
    DCT (CR);
    quantization (CR, ICqt);
    output_ptr = huffman (jpeg_encoder_structure, 3, output_ptr);
  return output_ptr;
}
UINT32 encode_image (UINT8 * input_ptr, UINT8 * output_ptr,UINT32 quality_factor, UINT32 image_format,UINT32 image_width, UINT32 image_height)
{
  UINT16 i, j;
  UINT8 * output;
  JPEG_ENCODER_STRUCTURE JpegStruct;
  JPEG_ENCODER_STRUCTURE * jpeg_encoder_structure = &JpegStruct;
  output = output_ptr;
image_format = FOUR_TWO_TWO;
RGB_2_422 (input_ptr, output_ptr, image_width, image_height);
/* Initialization of JPEG control structure */
initialization (jpeg_encoder_structure, image_format, image_width,
image_height);
/* Quantization Table Initialization */
initialize_quantization_tables (quality_factor);
/* Writing Marker Data */
output_ptr =
write_markers (output_ptr, image_format, image_width, image_height);
for (i = 1; i <= jpeg_encoder_structure->vertical_mcus; i++)
{
  if (i < jpeg_encoder_structure->vertical_mcus)
  jpeg_encoder_structure->rows = jpeg_encoder_structure->mcu_height;
  else
  jpeg_encoder_structure->rows =
  jpeg_encoder_structure->rows_in_bottom_mcus;
  for (j = 1; j <= jpeg_encoder_structure->horizontal_mcus; j++)
  {
   if (j < jpeg_encoder_structure->horizontal_mcus)
   {
    jpeg_encoder_structure->cols =
    jpeg_encoder_structure->mcu_width;
    jpeg_encoder_structure->incr =
    jpeg_encoder_structure->length_minus_mcu_width;
   }
   else
   {
    jpeg_encoder_structure->cols =
    jpeg_encoder_structure->cols_in_right_mcus;
    jpeg_encoder_structure->incr =
    jpeg_encoder_structure->length_minus_width;
   }
   read_format (jpeg_encoder_structure, input_ptr);
   /* Encode the data in MCU */
   output_ptr =encodeMCU (jpeg_encoder_structure, image_format, output_ptr);
   input_ptr += jpeg_encoder_structure->mcu_width_size;
  }
  input_ptr += jpeg_encoder_structure->offset;
}
/* Close Routine */
output_ptr = close_bitstream (output_ptr);
return (UINT32) (output_ptr - output);
}
void main()
{
int size=1280*720*3,sizeout=0,x,y;
unsigned char *outpict ;
outpict = (unsigned char *) malloc (size);
unsigned char *inpict ;
inpict = (unsigned char *) malloc (size);
for(x=0;x<1280;x++)//inpict初始化成RGB模式
  for (y=0;y<720;y++)
  {
   *(inpict+x*3+y*1280*3)=y%2;
   *(inpict+x*3+y*1280*3+1)=x%2;
   *(inpict+x*3+y*1280*3+2)=y%2;
  }
sizeout = encode_image (inpict,outpict,1,0,1280,720);
free(outpict);
free(inpict);
return 1;
}
(dog0138)

使用特权

评论回复 赏
板凳
ooppllmm| | 2014-1-10 17:41 | 只看该作者
不明觉厉

使用特权

评论回复 赏
地板
haso2007| | 2014-2-17 14:30 | 只看该作者
顶一个，留以后用。

使用特权

评论回复 赏
5楼
linhaishi| | 2014-2-20 09:08 | 只看该作者

使用特权

评论回复 赏
6楼
linzhiqi1| | 2014-3-31 21:31 | 只看该作者
bmp to jpg 吗

使用特权

评论回复 赏
7楼
PIAOYU| | 2014-5-14 00:37 | 只看该作者
挺好的，先收藏一下

使用特权

评论回复 赏
8楼
chddx222| | 2014-5-17 16:19 | 只看该作者
收藏

使用特权

评论回复 赏
9楼
那就地方iv| | 2015-8-15 22:12 | 只看该作者
谢谢楼主分享

使用特权

评论回复 赏
返回列表
发新帖 我要提问
高级模式
B Color Image Link Quote Code 收费 Smilies
您需要登录后才可以回帖 登录 | 注册

本版积分规则

small泡泡

3

主题

23

帖子

0

粉丝