mudhorn/zxing
1
0
Fork 0
mirror of https://github.com/zxing/zxing.git synced 2025-02-02 05:41:08 -08:00
Code Issues Actions Packages Projects Releases Wiki Activity
342 commits 6 branches 32 tags 331 MiB
Commit graph

187 commits

Author SHA1 Message Date
srowen 15f34edb52 Simple check in Code 39 reader to determine wide bars vary in width a lot, to cut out most false positives 
git-svn-id: https://zxing.googlecode.com/svn/trunk@442 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-18 22:52:45 +00:00
srowen 180e833b3e Improved approach to 1D decoding -- better use of integer math by scaling pattern ratios up to expected number of pixels, rather than the other way. Modified constants accordingly. Also introduced notion of maxium variance that any one bar in a pattern can have and stiill be accepted. Finally, adjusted false-positives test failure limit downward due to recent improvements. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@441 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-18 22:12:59 +00:00
srowen 2f9e5ca32b Fix pretty clear bug on Code 128 and 39 readers: reject empty barcodes as a false positive. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@440 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-18 21:56:30 +00:00
srowen 593128df02 Make optimization of core.jar as a library more conservative -- no shrinking, obfuscation 
git-svn-id: https://zxing.googlecode.com/svn/trunk@439 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-18 21:44:09 +00:00
srowen 6effdd4bd0 Renamed core OSGi bundle to "ZXing"; removed "level=" attribute on <jar> tasks since it singlehandly introduces a depence on Ant 1.7 and doesn't buy much. OS X for example does not ship with Ant 1.7 yet. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@437 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-18 17:40:14 +00:00
dswitkin 4d1229bd65 Created a library of images which do not contain barcodes, or contain barcodes we don't yet support. Wrote a new unit test designed to identify false positives, and set a benchmark which we can improve over time. Out of these 26 images, tested in all four rotations, we currently find 44/104 false positives. The Code 39 decoder in particular seems to be too lenient. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@436 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-17 18:35:16 +00:00
srowen c29b094932 Add support for core.jar as OSGi bundle, component in BugLabs's BUG platform 
git-svn-id: https://zxing.googlecode.com/svn/trunk@435 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-16 18:25:56 +00:00
srowen b5b4351727 Final updates to use local test images rather than remote 
git-svn-id: https://zxing.googlecode.com/svn/trunk@432 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-16 16:29:48 +00:00
srowen ab3a9bf493 More updates to use local test images rather than remote 
git-svn-id: https://zxing.googlecode.com/svn/trunk@431 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-16 16:13:12 +00:00
srowen f66a048f5c First set of updates to use local test images rather than remote 
git-svn-id: https://zxing.googlecode.com/svn/trunk@430 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-16 16:00:51 +00:00
srowen 70c64558cf Minor style tweaks 
git-svn-id: https://zxing.googlecode.com/svn/trunk@428 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-13 12:15:49 +00:00
srowen 6f7a348f5d Adjust optimization options: make sure to actually exclude more unused code, but keep what Android client needs; this means tests have to run against unoptimized version 
git-svn-id: https://zxing.googlecode.com/svn/trunk@427 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-12 17:11:07 +00:00
srowen 4cc490db56 Remove compile-time reference to DataMatrixReader 
git-svn-id: https://zxing.googlecode.com/svn/trunk@426 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-12 16:30:01 +00:00
srowen 56c4b6a94d Added build-optimized target to produce optimized core.jar; tests now test optimized version; also removed the dependency from core to MIDP 2.0 as it was just depending on silly IllegalStateException 
git-svn-id: https://zxing.googlecode.com/svn/trunk@423 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-12 15:55:42 +00:00
srowen 14895cf9b3 Now use integer math in patternMatchVariance for performance, plus a few other similar style and speed tweaks 
git-svn-id: https://zxing.googlecode.com/svn/trunk@422 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-12 15:17:53 +00:00
srowen 038add9da1 Oops one more style change 
git-svn-id: https://zxing.googlecode.com/svn/trunk@421 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-12 12:52:18 +00:00
srowen 31b8b3d226 Small tweaks on top of Daniel's excellent refactoring 
git-svn-id: https://zxing.googlecode.com/svn/trunk@420 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-12 12:51:26 +00:00
dswitkin 7b2cafe970 Missed one place that needs to cache the luminance data. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@419 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-11 23:58:04 +00:00
dswitkin 9adb83bbd3 Refactored the MonochromeBitmapSource hierarchy to share a great deal of code and shrink the derived classes considerably. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@418 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-11 23:52:35 +00:00
srowen f5b58916c1 Fix bad logic black point estimator, improving threshold estimation performance (and adjust some == to .equals()) 
git-svn-id: https://zxing.googlecode.com/svn/trunk@417 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-11 19:59:38 +00:00
srowen 88afca0842 Unify UPC-A reader into EAN-13 reader since EAN-13 is a superset and we can take advantage of the commonality to improve performance. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@416 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-11 19:30:57 +00:00
dswitkin f431cae59f - Added support for rotation in our blackbox test framework, and refactored the ways tests are created and run. 
- Turned on 0 and 180 degree rotation for all 1D formats.
- Turned on 0, 90, 180, and 270 degree rotation for QR.
- Changed the 1D code to re-enable upside down scanning, with a dramatic improvement in barcodes found and fewer false positives.

git-svn-id: https://zxing.googlecode.com/svn/trunk@411 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-11 00:05:51 +00:00
srowen 2cc9d7f630 Tiny bug fix, need to deal with no query params case correctly 
git-svn-id: https://zxing.googlecode.com/svn/trunk@410 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-06-10 19:44:47 +00:00
srowen a73fc5919a Added SMSTO support, added subject/body param support in mailto:, improved handling of mailto: and tel: URIs 
git-svn-id: https://zxing.googlecode.com/svn/trunk@402 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-23 22:40:13 +00:00
srowen c2c8b2fcfb Added handy toString() methods 
git-svn-id: https://zxing.googlecode.com/svn/trunk@401 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-23 22:20:10 +00:00
srowen 8d31073968 Detector is now a little more skeptical once it has found 3 confirmed finder patterns -- one may be a false positive, so it also checks to see if the estimated module sizes are "pretty similar". If not, keeps looking. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@394 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-08 15:12:44 +00:00
srowen 40f2cf1a80 Fix some minor javadoc issues 
git-svn-id: https://zxing.googlecode.com/svn/trunk@388 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-05 17:56:47 +00:00
srowen 08c7c6fadb Checking in somewhat ugly 'fix' for ArrayIndexOutOfBoundsException in rare cases 
git-svn-id: https://zxing.googlecode.com/svn/trunk@385 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-05 17:31:31 +00:00
srowen 635d904283 Many more little tweaks from IntelliJ inspections 
git-svn-id: https://zxing.googlecode.com/svn/trunk@382 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-02 19:18:14 +00:00
srowen 188c7fbecb Added preliminary support for MobileTag formats; not enabled yet. 
Added "sms:" URI support.

git-svn-id: https://zxing.googlecode.com/svn/trunk@380 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-05-02 18:39:03 +00:00
dswitkin 0987663963 Added some simple sharpening for 1D decoding which allowed a couple more blackbox images to pass. There were a few cases where the format or content is now misdetected, but since the net gain was positive I decided to make those non-fatal errors. In real world use the sharpening seems to help, and I think we can do even better with a better algorithm. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@372 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-21 18:11:02 +00:00
dswitkin a69a87e98d Replaced the smaller, resized EAN13-2 test images with the full-size originals. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@370 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-19 18:38:52 +00:00
srowen df25fc1389 Separate out test-unit and test-blackbox targets for running different sorts of tests 
git-svn-id: https://zxing.googlecode.com/svn/trunk@369 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-18 17:03:04 +00:00
dswitkin 9eceb0593e Added a new test library of 16 blurry EAN-13 barcodes. Currently we cannot decode any of them. I'm checking these in as a baseline so we can evaluate various sharpening techniques. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@368 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-18 15:51:39 +00:00
dswitkin a2662ebb40 Changed the black box tests to require an absolute number of images to decode successfully, instead of a percentage, and set the high water mark for each test. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@367 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-18 14:43:56 +00:00
dswitkin a816d5d9b1 Changed the parsing logic to refer to EAN-8 and EAN-13 as UPCs as well, so that the clients can recognize them as product barcodes and offer to do a product search. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@366 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-18 14:30:02 +00:00
srowen 989dc395f0 Added workaround for NoSuchMethodError on some Nokia JVMs with a superfluous abstract method (thanks Marnix) 
git-svn-id: https://zxing.googlecode.com/svn/trunk@364 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-16 00:38:54 +00:00
dswitkin ae54d7a937 Changed how many rows we scan while detecting QR finder patterns. The mobile case dynamically calculates this based on the height of the image, and ranges from a small speedup to a big difference for large inputs. Also changed the try harder case to scan every 3rd line for a 3x speedup. The unit tests have the same success rate as before: 16/20 on qr1, 10/10 on qr2. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@354 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-10 20:26:42 +00:00
srowen 1c9b906442 Workaround for codes that fail to include (required) final TERMINATOR mode indicator 
git-svn-id: https://zxing.googlecode.com/svn/trunk@351 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-07 22:05:40 +00:00
srowen a5b385e764 Standardize array initializer syntax to use the form without "new type[]", to be consistent and make translation to C++ easier 
git-svn-id: https://zxing.googlecode.com/svn/trunk@348 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-07 19:42:35 +00:00
srowen 9c1b4a1393 Make two methods public for easier reuse by third-party implementaitons 
git-svn-id: https://zxing.googlecode.com/svn/trunk@345 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-04 16:34:42 +00:00
srowen 5f25bd00a3 Moved the "geo:" fix to the 'right' place 
git-svn-id: https://zxing.googlecode.com/svn/trunk@343 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-03 22:18:16 +00:00
mschulkind 87da7e9226 small typo fix. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@341 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-03 20:34:30 +00:00
srowen d1d7e96238 At last, removing SKIP_N_BARCODES and separating this logic out in a way that individual projects can implement it on their own. It's a bit too onerous and project-specific to live on in the main code. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@340 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-03 16:39:44 +00:00
srowen 28ea8e02f0 Parsing wasn't correctly ignoring the ':' in prefix! 
git-svn-id: https://zxing.googlecode.com/svn/trunk@339 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-03 16:12:08 +00:00
srowen 0e7872020c Another fix to ensure that 2 barcodes with the same info are counted separately. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@338 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-03 15:39:06 +00:00
srowen 01075819c5 "Skip 0 barcodes" was incorrectly not returning the first barcode found 
git-svn-id: https://zxing.googlecode.com/svn/trunk@337 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-02 14:03:02 +00:00
srowen 012a8c0ce7 Fixed same ResultPoint bug in Code 128 and Code 39 reader; added convenient toString() to Result and GenericResultPoint; added arg checking for Result constructor too 
git-svn-id: https://zxing.googlecode.com/svn/trunk@336 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-02 13:40:29 +00:00
dswitkin f299f413e0 Fixed the result points for 1D decoding so that we once again get a green line in the Android client across the successful row. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@335 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-04-01 21:23:14 +00:00
srowen 11fc234d76 Completed basic support for NFC / NDEF formats applicable to 2D barcodes. Not yet enabled. 
git-svn-id: https://zxing.googlecode.com/svn/trunk@329 59b500cc-1b3d-0410-9834-0bbf25fbcc57
 2008-03-31 20:51:24 +00:00