Our company has a store with products. Approx 6-700 hundred of them. Some of the article numbers has some similarity, for example there can be one article 1300538 and one 130538. This is ugly and pepole make mistakes because of this.
If you can validate that you picked the right product by taking a picture of the front, where the article number is (and sometimes a barcode) and validate it against the order being served. You should be able to get the human errors down to a minimum...right?
I have tried a little with barcodes. But have not been able to read many of my picture files. And it got me thinking that maybe it is to complicated to read when the image can be quite different with different supliers.
Love to get some input. How Sensitive is the barcode reader? Perhaps im just incompetent
Please attach one image for investigation. I will give you feedback, you should be able to expect more than 99% of accuracy if you correctly use our barcode writer/reader.
I have provided a few samples for you, images marked "No" the program does not find any bar codes in. This are just samples, the only shoe carton is the only one thats close to "real".
When i run the "Yes_two.jpg" image. It found two bar codes. But i can just see one in the image.
If its supposed to be like 99% accurate. Then im making something wrong
First JPEG compression is just a bad option for OCR, OMR and barcode recognition. This is a lossless encoding scheme which create a lot of artifacts around characters & lines...
Then for your images:
The first 2 ones: They have not very good contrast. So you should threshold them using mid-level linear transformation. IE: object.ConvertTo1Bpp(m_ImageID, 125).
The third: The barcode is invalid or the encoding scheme is not supported by GdPicture. See supported type from http://www.gdpicture.com/products/plugi ... nition.php
Hope this helps !
Who is online
Users browsing this forum: No registered users and 3 guests