How to enable OCR with Scangear MP under Linux for MG8250?

[vlenhardt]

Under Linux openSUSE 12.3 I use Scangear MP 1.80 for my Canon PIXMA MG8250.

As a scan target (in my German GUI it's "Ziel") I can choose "OCR". But there is no text output, the scan is just saved as a picture. How can I configure Scangear to open the scan via Tesseract in a text editor.

By the way, the target "Print" (in German "Drucken") leads to saving just as well, though there is at least one printer at hand, of course.

It's a pity that the pixma scan drivers that are shipped with openSUSE don't work with xsane (the scanner starts working, but stops after one third of the page). Using xsane with a different scanner I can use OCR successfully.

Volker

[Stephen]

Hi vlenhardt!

Welcome to The Canon Forums, and thank you for your inquiry! We appreciate your participation, however we need to let you know that your product seems to be a model that is not supported by our team here at Canon USA. The Canon Forum is hosted and moderated within the United States by Canon USA. We are only able to provide support for Canon products manufactured for the US market. If you live outside the United States, please click here and select your country or region for your support needs. Feel free to discuss Canon products sold outside of the United States, but please be aware that you will not receive support directly from Canon USA.

[vlenhardt]

Thank you for the link, Stephen. I've just posted a support request.

The reason for asking my question here is that this is the only forum I could find to discuss these problems. I also was ensnared by the first welcoming sentence in the forum: "The Canon Forum is a convenient, interactive community where you can learn about Canon products, ask questions, or share your knowledge with your peers."

So I didn't mean to get support, but to get good advice. Sorry that I was wrong.

Volker

[arronlee]

There are two basic types of core OCR algorithm, which may produce a ranked list of candidate characters.
Matrix matching involves comparing an image to a stored glyph on a pixel-by-pixel basis; it is also known as "pattern matching" or "pattern recognition". This relies on the input glyph being correctly isolated from the rest of the image, and on the stored glyph being in a similar font and at the same scale. This technique works best with typewritten text and does not work well when new fonts are encountered. This is the technique the early physical photocell-based OCR implemented, rather directly.
Feature extraction decomposes glyphs into "features" like lines, closed loops, line direction, and line intersections. These are compared with an abstract vector-like representation of a character, which might reduce to one or more glyph prototypes. General techniques of feature detection in computer vision are applicable to this type of OCR, which is commonly seen in "intelligent" handwriting recognition and indeed most modern OCR software. Nearest neighbour classifiers such as the k-nearest neighbors algorithm are used to compare image features with stored glyph features and choose the nearest match.
Software such as Cuneiform and Tesseract use a two-pass approach to character recognition. The second pass is known as "adaptive recognition" and uses the letter shapes recognized with high confidence on the first pass to better recognize the remaining letters on the second pass. This is advantageous for unusual fonts or low-quality scans where the font is distorted.

So I wonder which kind of OCR algorithm are you using?

[vlenhardt]

I use tesseract.

I was looking for a way to connect a script to the output of scangearmp to produce text on the fly. I find that with scangearmp you can choose between three different output targets: Print, Screen, and OCR (I don't know the actual wording, because I have the German localisation). But whichever I select, the scan is always saved as a file.

In the meantime I got the scanner working with sane, thus getting tesseract working with xsane, too.

Nevertheless I think that a program such as scangearmp could be made to do what it promises. Can there be any hope for improvement? Or is there a way to get to print, screen and OCR output I haven't detected yet? All I need is some config file to connect an application to the output.

Kind regards

Volker