文件列表:

test1\.tif (4200, 2003-01-08)
test1\0.tif (16720, 2003-01-13)
test1\0_Read_Me_First.txt (5567, 2006-07-16)
test1\1.tif (16570, 2003-01-08)
test1\2.tif (3878, 2003-01-08)
test1\3.tif (2863, 2003-01-08)
test1\4.tif (3547, 2003-01-08)
test1\5.tif (1996, 2003-01-08)
test1\6.TIF (3936, 2003-01-08)
test1\7.tif (3863, 2003-01-08)
test1\adpmedian.p (2900, 2006-07-15)
test1\average.p (694, 2006-07-15)
test1\baboon.BMP (786486, 2010-12-27)
test1\barbara.BMP (786486, 2010-12-27)
test1\bayesgauss.p (3537, 2006-07-15)
test1\bdct.m (614, 2003-10-30)
test1\bdctmtx.m (506, 2003-10-30)
test1\bound2eight.p (4181, 2006-07-15)
test1\bound2four.p (3929, 2006-07-15)
test1\bound2im.p (4307, 2006-07-15)
test1\boundaries.p (7577, 2006-07-15)
test1\bsubsamp.p (4403, 2006-07-15)
test1\cameramantest.asv (623, 2010-12-27)
test1\cameramantest.m (800, 2010-12-27)
test1\changeclass.p (958, 2006-07-15)
test1\channal_test.asv (604, 2011-06-16)
test1\channal_test.m (607, 2011-06-16)
test1\channel_test.asv (817, 2011-06-20)
test1\channel_test.m (817, 2011-06-20)
test1\colorgrad.p (4904, 2006-07-15)
test1\colorseg.p (3131, 2006-07-15)
test1\compare.p (1982, 2006-07-15)
test1\connectpoly.p (1494, 2006-07-15)
test1\Contents.m (6129, 2006-07-15)
test1\conwaylaws.p (1069, 2006-07-15)
test1\covmatrix.p (1152, 2006-07-15)
test1\dequantize.m (563, 2003-10-30)
test1\dftcorr.p (646, 2006-07-15)
test1\dftfilt.p (790, 2006-07-15)
test1\dftuv.p (1124, 2006-07-15)
... ...

Matlab JPEG Toolbox =================== This distribution contains routines for manipulating files formatted according to the Joint Photographic Experts Group (JPEG) standard. Matlab's built-in IMREAD and IMWRITE functions provide basic conversion between JPEG files and image arrays. The routines in this package provide additional functionality for directly accessing the contents of JPEG files from Matlab, including the Discrete Cosine Transform (DCT) coefficients, quantization tables, Huffman coding tables, color space information, and comment markers. It is assumed that the user of this software has a good understanding of both the JPEG compression standard and Matlab data structures. This software is based in part on the work of the Independent JPEG Group (IJG), as it makes use of IJG's free JPEG code library. If the MEX file binaries provided in this distribution are not the ones you need for your system, you will need to download IJG's code library and install it on your system before compiling the source code. See "Installing" for more details. If you find this software useful, or if you would like to contribute to the project, please send me email. Phil Sallee 9/2003 Copyright Notice ================ Copyright (c) 2003 The Regents of the University of California. All Rights Reserved. Permission to use, copy, modify, and distribute this software and its documentation for educational, research and non-profit purposes, without fee, and without a written agreement is hereby granted, provided that the above copyright notice, this paragraph and the following three paragraphs appear in all copies. Permission to incorporate this software into commercial products may be obtained by contacting the University of California. Contact Jo Clare Peterman, University of California, 428 Mrak Hall, Davis, CA, 95616. This software program and documentation are copyrighted by The Regents of the University of California. The software program and documentation are supplied "as is", without any accompanying services from The Regents. The Regents does not warrant that the operation of the program will be uninterrupted or error-free. The end-user understands that the program was developed for research purposes and is advised not to rely exclusively on the program for any reason. IN NO EVENT SHALL THE UNIVERSITY OF CALIFORNIA BE LIABLE TO ANY PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN IF THE UNIVERSITY OF CALIFORNIA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. THE UNIVERSITY OF CALIFORNIA SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND THE UNIVERSITY OF CALIFORNIA HAS NO OBLIGATIONS TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS. Installing ========== Copy all of the files into a new directory. Add this directory to the Matlab startup path, if desired. Compile the MEX routines, if necessary. Compiled MEX routines with the extension .dll, are provided for Windows 9x/NT/2000 systems. These were compiled for Matlab 6.5.0.180913a (R13) using the Microsoft Visual C++ compiler. If you have a different version of Matlab, they may not work on your system. For use on other platforms, you will need to take the following steps to compile the MEX routines. In order to compile the MEX routines, you must first build IJG's JPEG Tools code library, available at ftp://ftp.uu.net/graphics/jpeg/jpegsrc.v6b.tar.gz. Download and build the libjpeg library using the make files and instructions contained in the IJG JPEG distribution. Once the IJG libjpeg library has been built, execute the following commands from the command shell or from inside a Matlab window, depending on your Matlab configuration: mex -I jpeg_read.c mex -I jpeg_write.c Replace with the path to the IJG jpeg-6b directory, and with the full path to the IJG code library file, generally saved as libjpeg.a or libjpeg.lib depending on the operating system. Files contained in this distribution ==================================== README bdct.m bdctmtx.m dequantize.m ibdct.m im2vec.m jpeg_qtable.m jpeg_read.c jpeg_read.dll jpeg_read.m jpeg_write.c jpeg_write.dll jpeg_write.m quantize.m vec2im.m Matlab Functions ================ The following Matlab functions are included in this distribution: jpeg_read Read a JPEG file into a JPEG object struct jpeg_write Write a JPEG object struct to a JPEG file jpeg_qtable Generate standard JPEG quantization tables bdct Blocked discrete cosine transform ibdct Inverse blocked discrete cosine transform bdctmtx Blocked discrete cosine transform matrix (2D, full transform) quantize Quantize BDCT coefficients (using a quantization table) dequantize Dequantize BDCT coefficients, using center bin estimates im2vec Reshape 2D image blocks into an array of column vectors vec2im Reshape and combine column vectors into a 2D image The following routines may be added in a future version of this toolbox: jpeg_info Display pertinent information about a JPEGOBJ struct jpeg2im Convert a JPEGOBJ struct to an image im2jpeg Convert an IMAGE to a JPEGOBJ struct huff_encode Huffman encode coefficients into a bitstream huff_decode Huffman decode a bitstream into coefficients bitpack Repack array of bits into an array of bytes, or vice versa Release Notes ============= 9/08/03 v1.0 Initial release 9/24/03 v1.1 Fixed problem with bdct/ibdct and non-square images 9/25/03 v1.2 Corrected warnings in jpeg_read/jpeg_write Workaround for incompatibility with older Matlab versions 10/14/03 v1.3 Fixed seg faults caused by warnings in jpeg_read, jpeg_write Added error checking for quantization table entries Added force_baseline option to jpeg_qtable 10/30/03 v1.4 Added basic support for progressive mode JPEG Documentation ============= Type 'help ' for documentation on a specific function. At this point, relatively little documentation is provided. The tools are simple, however. The jpeg_read, and jpeg_write routines convert information from a JPEG file into a Matlab struct, which I will also refer to as a "jpeg object". This Matlab struct contains all of the critical information present in the JPEG file, in the form of arrays, cell arrays and nested structs indexed by (hopefully intuitive) field names. The only information present in the JPEG file not contained in the jpeg object are currently the Adobe markers. I may add that functionality in a future release. The field names for the structs generally match the names used in the JPEG standard and in the IJG code library. These were chosen to be as intuitive as possible. If there are multiple color components in the JPEG file, as is typical, the coefficient array for each component is stored in a cell array indexed by the component number. Information specific to the component is stored in the field comp_info, an array of structs indexed by the same component number. The quantization tables can be found in the field quant_tables, which contains a cell array of tables. The index of the quantization table used by for a given component is stored in the quant_tbl_no in the comp_info struct. For example, to read an image into a jpeg object named 'jobj'. >> jobj = jpeg_read('myimage.jpg'); List the contents of jobj (which is really a Matlab struct): >> jobj jobj = image_width: 227 image_height: 149 image_components: 0 image_color_space: 2 jpeg_components: 3 jpeg_color_space: 3 comments: {} coef_arrays: {[152x232 double] [80x120 double] [80x120 double]} quant_tables: {[8x8 double] [8x8 double]} ac_huff_tables: [1x2 struct] dc_huff_tables: [1x2 struct] optimize_coding: 0 comp_info: [1x3 struct] progressive_mode: 0 This gives us some basic information such as image size, number of color components, etc. To access the quantization table used by the first component: >> jobj.quant_tables{jobj.comp_info(1).quant_tbl_no} ans = 8 6 5 8 12 20 26 31 6 6 7 10 13 29 30 28 7 7 8 12 20 29 35 28 7 9 11 15 26 44 40 31 9 11 19 28 34 55 52 39 12 18 28 32 41 52 57 46 25 32 39 44 52 61 60 51 36 46 48 49 56 50 52 50 This corresponds to a luminance channel quantization table with a quality of 75, which you can check using jpeg_qtable: >> jpeg_qtable(75) ans = 8 6 5 8 12 20 26 31 6 6 7 10 13 29 30 28 7 7 8 12 20 29 35 28 7 9 11 15 26 44 40 31 9 11 19 28 34 55 52 39 12 18 28 32 41 52 57 46 25 32 39 44 52 61 60 51 36 46 48 49 56 50 52 50 Note: The indices for quant_tbl_no, ac_huff_table, and dc_huff_table stored in the actual JPEG file are 0-based, but in the jpeg object struct these are converted to be 1-based so that they correspond to Matlab's indexing convention for the cell arrays. All other values are directly copied to and from the JPEG file without conversion. Setting jobj.optimize_coding to 1 before calling jpeg_write will cause the compressor to optimize the huffman coding tables. This usually provides a small percentage decrease in file size. Note that this will case the huffman tables in the object to be ignored when the file is written. >> jobj.optimize_coding = 1; To add a comment to the jpeg object: >> jobj.comments = {'Comment: The huffman tables are optimized'}; Write out the jpeg file with the added comment, optimizing the huffman tables: >> jpeg_write(jobj,'myimage2.jpg'); If the file is read back into another jpeg object, the optimize_coding field will not be set to 1 in the new object. That is because this information is not contained within the JPEG file. It is merely a code to the compression engine to optimize the tables. A clever program could look at the tables and take a guess as to whether they have been optimized based on whether they match the standard tables, but there is no saved value in the JPEG file that indicates if the huffman tables were optimized. The JPEG colorspace constants (found in IJG's jpeglib.h) are: 0 JCS_UNKNOWN, /* error/unspecified */ 1 JCS_GRAYSCALE, /* monochrome */ 2 JCS_RGB, /* red/green/blue */ 3 JCS_YCbCr, /* Y/Cb/Cr (also known as YUV) */ 4 JCS_CMYK, /* C/M/Y/K */ 5 JCS_YCCK /* Y/Cb/Cr/K */ See the documentation in the IJG code library for more information on image/jpeg colorspaces. I plan to add a function jpeg_info at a later date which will interpret some of the the raw data in a jpeg object and display plain text descriptions of such things as the color space, compression quality, and whether the huffman tables are generic or optimized. The BDCT and IBDCT routines convert between a single channel image and the blocked DCT transform coefficients. This can be useful for interpreting the coefficient arrays. The QUANTIZE and DEQUANTIZE routines take a coefficient array and quantization table and convert between the coefficient values and the quantization indices. As of version 1.4 there is a field to indicate if the JPEG file was stored in progressive_mode. This flag can be modified to turn on/off progressive mode when writing to a file. Currently, the scan sequence is not preserved. If progressive mode is enabled, a default scan sequence is used. Also, turning on progressive mode results in using optimized huffman tables, as the default huffman tables are unsuitable for progressive files. More information about the JPEG compression standard can be found at the following reference: Wallace, Gregory K. "The JPEG Still Picture Compression Standard", Communications of the ACM, April 1991 (vol. 34 no. 4), pp. 30-44.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：