Need public function to tell kernel type used (5x3 vs 9x7)

[gcode-importer]

Originally reported on Google Code with ID 3

 AFAIK j2k.h and jp2.h header files are not installed on a standard
openjpeg installation. Is this intended ?
  Here is the issue I am currently facing, in order to reject lossy
JP2/J2K stream in clinical trial environment I had to code the
following:

  int reversible;
  opj_j2k_t* j2k = NULL;
  opj_jp2_t* jp2 = NULL;

  switch(parameters.decod_format)
    {
  case J2K_CFMT:
    j2k = (opj_j2k_t*)dinfo->j2k_handle;
    assert( j2k );
    reversible = j2k->cp->tcps->tccps->qmfbid;
    break;
  case JP2_CFMT:
    jp2 = (opj_jp2_t*)dinfo->jp2_handle;
    assert( jp2 );
    reversible = jp2->j2k->cp->tcps->tccps->qmfbid;
    break;
  default:
    gdcmErrorMacro( "Impossible happen" );
    return false;
    }
  LossyFlag = !reversible;

Unfortunately I need to include j2k.h  / jp2.h which is completely
non-standard on linux distribution.

What other people recommend for this situation:

- Extend openjpeg installation mechanism to install j2k.h / jp2.h
- Extend openjpeg.h API to return the reversible flag.

Thanks for suggestion 

Reported by malaterre on 2009-06-08 08:33:41

[gcode-importer]

Ref:
http://groups.google.com/group/openjpeg/browse_thread/thread/7c80e293e80fa931

Reported by malaterre on 2009-06-08 08:33:52

[gcode-importer]

Issue is still current in 2012:
http://groups.google.com/group/openjpeg/browse_thread/thread/af13483b979b77b8

Reported by malaterre on 2012-01-11 19:11:31

• Status changed: Accepted
• Labels added: Priority-High
• Labels removed: Priority-Medium

[gcode-importer]

Following the norm, each tile-component can be coded with is own transformation (reversible
or irreversible) so is quite difficult to have one unique function: cf markers COD
and COC

Reported by savmickael on 2012-11-13 12:58:30

[gcode-importer]

Reported by malaterre on 2014-02-25 15:19:50

• Labels added: Priority-Low
• Labels removed: Priority-High

[gcode-importer]

Reported by malaterre on 2014-02-27 08:04:29

• Labels added: Milestone-Release2.1

[gcode-importer]

Reported by savmickael on 2014-03-06 11:43:54

[gcode-importer]

Creating a fake irreversible image:

$ kdu_compress -no_info -i lena512.pgm -o lena512.j2k Ckernels:T0C0=W5X3

Reported by malaterre on 2014-03-14 13:11:34

[gcode-importer]

This issue was updated by revision r2740.

Reported by malaterre on 2014-03-14 13:53:47

[gcode-importer]

Funky one:
$ kdu_compress -no_info -i lena512.pgm -o lena512.j2k Ckernels:T2C0=W9X7 "Stiles={256,256}"
Ckernels:T1C0=W5X3

Reported by malaterre on 2014-03-14 13:54:29

[gcode-importer]

More fun:
$ kdu_compress -no_info -i lena512.pgm -o lena512.j2k Ckernels:T2C0=W9X7 "Stiles={256,256}"
Ckernels:T1C0=W5X3 Clevels:T1C0=0 Cuse_sop:T2=yes Cprecincts:T0="{128,128}"

Reported by malaterre on 2014-03-14 14:00:17

[gcode-importer]

Reported by malaterre on 2014-03-14 14:02:38

• Labels removed: Milestone-Release2.1

[malaterre]

I used to have the following code on my local openjpeg version in GDCM:

int j2k_get_reversible(
  opj_j2k_t * p_j2k)
{
  opj_cp_t *cp = 00;
  cp = &(p_j2k->m_cp);
  return cp->tcps->tccps->qmfbid;
}
int jp2_get_reversible(
  opj_jp2_t * p_jp2)
{
  return j2k_get_reversible(p_jp2->j2k);
}

int OPJ_CALLCONV opj_get_reversible(opj_codec_t *p_info, opj_dparameters_t *parameters)
{
  int ret = -1;
  if (p_info)
    {
    opj_codec_private_t * l_info = (opj_codec_private_t *) p_info;
    if (l_info->is_decompressor)
      {
      switch(parameters->decod_format)
        {
      case 0: // J2K_CFMT:
        ret = j2k_get_reversible(l_info->m_codec);
        break;
      case 1: // JP2_CFMT:
        ret = jp2_get_reversible(l_info->m_codec);
        break;
        }
      }
    }
  return ret;
}

[malaterre]

One may need to adapt the suggested code to loop over all tiles.

[malaterre]

Not having such functions prevents me from switching from OPJ 1.5 to OPJ 2.1 see ref: https://sourceforge.net/p/gdcm/bugs/356/

[malaterre]

The longer story is that (usual UNIX stuff) programs are now being compiled against opj 1.5 and somewhere deep down in the dep chains, another lib is compiled against opj 2.1 (different ABI). This may become a very serious issue in the not so distant future.

[malaterre]

As a side note, I would also be required to expose the MCT (COD) flag also. This helps DICOM encapsulation where knowing the transformation is required.

[rouault]

This information is available through the use of opj_get_cstr_info(). Admitedly this is just for the default coding parameters, not per-tile. But deciding the filter per-tile is rather unusual