1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
|
0.30: (2024-12-24)
* Reject odd input depths that pixThresholdToBinary doesn't support.
* Fix build with Leptonica >=1.83
* Replace vs2008 solution with CMake
* Fix `Error heap-use-after-free`
* Fix typos in code
* Support python 3, drop support for python2
* Rename pdf.py to jbig2topdf.py
* Add a A switch to set dpi for images w/o dpi information
* Add support for standalone .jb2
* Suppress a chatty informational message unless requested
* Neat symbolic threshold 0.92
* Fix Endian issue on ARM #63
* Neat symbolic threshold 0.92
* Make the weight parameter adjustable
* Add lobal BW threshold on 8 bpp images
* Replace obsolete macro `AC_PROG_LIBTOOL'
0.29 (2017-01-30)
* fix build with leptonica>=1.70
* Remove incorrect reference count update.
0.28 (2012-09-19):
* Update to the latest Leptonica (1.68)
* autotools support, VC++ 2008 solution
* fix binary file open mode on Windows
* version info (-V --version)
* pdf.py now correctly retains DPI from input images (thanks to Steven Lee
http://blog.rubypdf.com/2011/09/09/jbig2-pdf-py-patch-the-right-way-to-get-dpi/)
* R. Hatlapatka: option to use autoThresholding. Improved version from
bachelor thesis JBIG2 compression http://is.muni.cz/th/208155/fi_b/.
0.27 (2009-04-29):
* Update to the latest Leptonica (1.58)
* Fix comments in pdf.py (thanks Dan)
0.26:
* Update to the latest Leptonica (1.53)
0.25:
* Should build on Windows if you get a leptonica which does so (thanks to
Dwight Kelly and a man known only as Dennis)
0.24:
* Update to leptonica 1.38
* Add ability to gray symbols but it's commented out because it does little
good
* Refinement support broke because of memory savings in leptonica
0.23:
* Quick release to support leptonica 1.37 (now bundled)
0.22:
* Added segmentation support to extract text from mixed images
* Flyspeck removal disabled for now
* A few minor fixes
0.21:
* Patch from Alberto Accomazzi to fix the pdf.py script and to output more
sensible filenames in pdf mode
* Symbols which are only used once are now included in a per-page dictionary
(great for scans of old books)
* Fixed several scale bugs (one at 65K symbols and one at 100K symbols)
* Page numbers > 256 went wrong
|
