You are viewing limited content. For full access, please sign in.

Question

Question

Barcode Issues with QF - confused....

asked on November 23, 2016

Hi All

 

See the attached screen shot, we are bring a new invoicing system online and all invoices are being scanned into LF via QF (v10).

 

We scan the barcode, w/checksum and then set the filename to the invoice number - simple as that. However on about 1-2% of the scans we are getting invalid results from the barcode scan.  Like bad OCR.  This is a typical example - scanning the page directly works as expected.

 

How does a barcode w/checksum scan bad?

 

Any ideas?

 

 

Jeff

 

 

PGC-LFSWORK_USscanQuestions.jpg
0 0

Replies

replied on November 23, 2016 Show version history

Just like with OCRing text, poor scan quality (or print quality of the original) can cause issues misreading barcodes when QF can't tell the difference between, for example, a double-width and a single-width line. In a previous discussion on the same topic, the QF team recommended trying the "smooth" process to clean up the lines, and another customer recommended cleaning or tuning of the scanner.

The patterns of the lines and spaces turn into binary representations specific characters (see here for an example of the encoding), so very minor differences lead to a completely different letter, number, or symbol. 1101101100 is the symbol !, but 11011001100 is the number 0.

For example (with extra spacing for clarity):

1 0
replied on November 24, 2016

Thanks for the info, we are using check digits on our barcodes - with that in mind wouldn't a failed check digit check cause a failed scan?

 

With that in mind how would I ever get a H!000808AS type result?   Now it is possible that our problem is the new invoicing system, perhaps they are not actually using a check digit, even though it appears that they are.

 

 

0 0
replied on November 28, 2016 Show version history

Quick Fields will just return a blank value for a barcode that fails the checksum, to handle 'false positives' where there is not actually a barcode there at all. I noticed your sample page has multiple barcodes - is it possible that a different barcode is being read instead of the one you intended? 

 

Edit: Actually, after looking closer at what you got vs. the format of the other barcodes, I'd guess that the image quality issues as Brianna's post describes is the source of the problem after all - especially since the image in your screenshot looks pretty rough. Even if you are using checksums, it's possible that either (a) multiple misread characters within the same barcode cancelled each other out in such a way that the checksum happened to still be correct or (b) that the checksum digit itself was misread. 

1 0
You are not allowed to follow up in this post.

Sign in to reply to this post.