That is the conclusion I have to come to as well. This may still work if one of our vendors adds an enhancement to their product but I will not hold my breath.
Thanks Gord From: Allison, Timothy B. [mailto:talli...@mitre.org] Sent: July 21, 2016 11:57 AM To: user@tika.apache.org Subject: RE: Extract Text from a TIFF image There are commercial OCR options that might fit the bill, but I'm not aware of any good open source pure Java OCR options. From: Gordon Schneider [mailto:schneid...@transampiping.com] Sent: Thursday, July 21, 2016 12:26 PM To: 'user@tika.apache.org' <user@tika.apache.org<mailto:user@tika.apache.org>> Subject: RE: Extract Text from a TIFF image I downloaded the version you mentioned and I got it to work. The results were the same as yours. But after review this unfortunately will not work for us as we need something that is in a jar file that we can install on our Power Server. The project that the OCR will eventually be part has this as a requirement. If you have any other suggestions, please let me know. Thanks for your time. It is very much appreciated. Gord From: Allison, Timothy B. [mailto:talli...@mitre.org] Sent: July 19, 2016 9:58 AM To: user@tika.apache.org<mailto:user@tika.apache.org> Subject: RE: Extract Text from a TIFF image You might want to experiment with different -psm values, we use 1 by default. Also, which version of Tesseract? I think I got mine from (https://github.com/UB-Mannheim/tesseract/wiki), version: tesseract 3.05.00dev leptonica-1.73 libgif 4.1.6(?) : libjpeg 8d (libjpeg-turbo 1.4.2) : libpng 1.6.20 : libtiff 4.0.6 : zlib 1.2.8 : libwebp 0.4.3 : libopenjp2 2.1.0 From: Gordon Schneider [mailto:schneid...@transampiping.com] Sent: Tuesday, July 19, 2016 11:22 AM To: 'user@tika.apache.org' <user@tika.apache.org<mailto:user@tika.apache.org>> Subject: RE: Extract Text from a TIFF image I installed tesseract on my PC. I ran tesseract on its own using the following command: tesseract.exe x:/java/PDFBox/Maxfield-1.tiff x:/java/PDFBox/Maxfield-1 The results are in the attached file. Not as clean as the results Timothy got. I am closer to where I want to get to but obviously I am a number of steps to my ideal solution. How to get the same results Timothy got? Thanks Gord From: Allison, Timothy B. [mailto:talli...@mitre.org] Sent: July 18, 2016 2:25 PM To: user@tika.apache.org<mailto:user@tika.apache.org> Subject: RE: Extract Text from a TIFF image You'll need to set up tesseract to run Optical Character Recognition. While we have an integration with OCR, it is not bundled within the app. See https://wiki.apache.org/tika/TikaOCR For kicks, I ran this through Tika+Tesseract; this is the output you get once you've set up Tesseract: SUPPLIER: 3177 Invoice Date Description Amount Discount Net Amount 015-28339 06/08/2015 21,318.54 0.00 21,318.54 C15-28837 06/04/2015 1,529.75 0.00 1,529.75 01528978 06/04/2015 1,238.18 0.00 1,238.18 015-28978-01 06/04/2015 1,182.85 0.00 1,182.85 015-28439 06/01/2015 1,113.86 0.00 1,113.86 C15-29707 06/11/2015 886.84 0.00 886.64 C15-28978-02 06/04/2015 526.91 0.00 526.91 01529385 06/09/2015 199.29 0.00 199.29 C15~28439~01 06/03/2015 157.34 0.00 157.34 C15-28670 06/03/2015 136.52 0.00 136.52 C15-28314-01 06/03/2015 132.81 0.00 132.81 015-28576 06/02/2015 61.26 0.00 61.26 015-29413 06/11/2015 22.37 0.00 22.37 Cheque #: 83077 Cheque Date 7/14/2015 28,506.32 0.00 28,506.32 SUPPLIER: 3177 Invoice Date Description Amount Discount Net Amount C15-28339 06/08/2015 21,318.54 0.00 21,318.54 015-28837 06/04/2015 1,529.75 0.00 1,529.75 015-28978 06/04/2015 1,238.18 0.00 1,238.18 015-28978-01 06I04/2015 1 ,18285 0.00 1,182.85 C15-28439 06/01/2015 1,113.86 0.00 1,113.86 015-29707 06l11/2015 886.64 0.00 886.64 C15-28978~02 06/04/2015 526.91 0.00 526.91 015-29385 06/09/2015 199.29 0.00 199.29 C15-28439-01 06/03/2015 157.34 0.00 157.34 015-28670 06/03/2015 136.52 0.00 136.52 015-28314-01 06/03/2015 132.81 0.00 132.81 C15-28576 06/02/2015 61.26 0.00 61.26 015-29413 06/11/2015 22.37 0.00 22.37 Cheque #1 83077 Check Daie: 7/14/2015 28,506.32 0.00 28,506.32 07142015 MMDDYYYY TWENTY-EIGHT THOUSAND FIVE HUNDRED SIX CAD AND 32/ 100 $ "******28,506.32 Trans Am Piping Canada From: Gordon Schneider [mailto:schneid...@transampiping.com] Sent: Monday, July 18, 2016 4:05 PM To: 'user@tika.apache.org' <user@tika.apache.org<mailto:user@tika.apache.org>> Subject: Extract Text from a TIFF image I have tried using the GUI for tika-app-1.13 but it shows nothing. I can see the metdata but that does not give me the information I need. I have attached the file. Maybe it is not possible to extract the text. If so what should I be looking for to tell me that it cannot extract the text. Thanks Gordon Schneider 403-236-0601 Trans Am Piping Products Ltd.
