sp_code.doc

常用的无损压缩算法源代码

 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

      S + P   I M A G E   C O M P R E S S I O N   P R O G R A M S

 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

               < ANSI C > version  4.03  -  05/30/95

       Amir Said: [e-mail] - [e-mail]
       DENSIS - Faculty of Electrical Enginnering
       UNICAMP - Campinas, SP, Brazil

       William A. Pearlman - [e-mail]
       Dept. of Electrical, Computer, and Systems Engineering
       Rensselaer Polytechnic Institute - Troy, NY 12180, USA


 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

  Copyright (c) 1995 Amir Said & William A. Pearlmam

  These programs are Copyright (c) by Amir Said and William A. Pearlman.
  They may be freely  redistributed in their entirety provided that
  this copyright notice is not removed. They may not be sold for profit
  or incorporated in commercial programs without the written permission
  of the copyright holders. Each program is provided as is, without any
  express or implied warranty, without even the warranty of fitness for
  a particular purpose.

 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



DESCRIPTION
===========

The programs "sp_compress", "sp_uncompress", "sp_pack", and "sp_unpack"
contain implementations of lossless image compression algorithms which
use the S+P transform , introduced in the paper, "Image Compression via 
Multiresolution Representation and Predictive Coding," by A. Said and 
W. A. Pearlman, in Visual Communications and Image Processing '93, 
Proc. SPIE 2094, pp. 664-674, Nov. 1993.  These programs compress gray-
scale images and decompress them to perfect recovery. 

The programs "sp_compress" and "sp_uncompress" use an arithmetic code
and yield the maximum compression. The programs "sp_pack" and "sp_unpack"
use a Huffman code, and yield smaller compression but with a smaller
compression/recovery time.

The arithmetic code implementation is in the files "arithm.h" and
"arithm.c", and is described in the file "arithm.doc". The Huffman
code is implemented and described in the files "huffman.h", "huffman.c",
and "huffman.doc".

Please send corrections, suggestions and bug reports to 
[e-mail].


FILES
=====

The programs are divided in several files, which are shared by the different
versions. To be included by the compiler they should be in the same directory.
The list of files is:

     - arithm.h           - arithm.c           - arithm.doc
     - compile.bat        - huffman.h          - huffman.c
     - huffman.doc        - sp_code.c          - sp_code.doc
     - sp_comm.c          - sp_compress.c      - sp_decd.c
     - sp_pack.c          - sp_uncompress.c    - sp_unpack.c



COMPILATION
===========

To compile these programs under UNIX, just type, at the propt

  compile.bat COMP

where 'COMP' is the name of the desired ANSI C compiler name (e.g.
COMP=gcc for GNU's C compiler, COMP=CC for Sun's, and
COMP=xlc for IBM's.


USAGE
=====

The original images should be in raster format, but without the header
(they are here called PIC image files).

The Huffman compression program usage is (pel = pixel):

sp_pack [-options] hor_pels [vert_pels] bytes/pel PIC_file compressed_file

The valid options are:

  -s           use for very smooth (usually medical) images
  -r           rectangular image (default = square)
  -sr, -rs     combination of the options above

It is assumed that the image is stored in a sequence of horizontal lines
with "hor_pels" pixels. If the image is square (default) then the parameter
"[vert_pels]" with the number of lines should not be included.

Most images use 1 byte/pixel (256 gray levels). Some medical images
use 2048 or more gray levels and are stored with 2 bytes/pixel. Images
processed with FORTRAN programs, stored as short integers also use
2 bytes/pixel. The most significant byte is read first.

The command to recover the original image is:

sp_unpack  compressed_file  PIC_file

Similarly, the usage for arithmetic compression is:

sp_compress [-options] [vert_pels] hor_pels bytes/pel PIC_file comp_file
sp_uncompress  compressed_file  PIC_file

When called without arguments these programs show the usage message.


EXAMPLES
========

sp_compress 512 1 lena.pic lena.spc
sp_uncompress lena.spc lena.pic

sp_pack -s 512 2 brainscan.dat brainscan.spp
sp_unpack brainscan.spp brainscan.dat

sp_pack -rs 1024 768 2 x_ray.pic x_ray.spp
sp_unpack x_ray.spp new_x_ray.pic


REMARKS
=======

-> These programs were tested in several images and compiled with the
   GNU, SUN, IBM, Borland and Microsoft C compilers. Nevertheless,
   they may not work with compilers that have other conventions for bit
   shifts of negative integers. Before deleting the original image it
   is always better to code and decode the image, and then compare the
   recovered version with the original. This can be done with the DOS 
   "fc" or the UNIX "cmp" commands.

 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -