Page 1 of 32 123411 ... LastLast
Results 1 to 15 of 470

Thread: RegulatedNoise - trade tool with zxctypo's best-in-class OCR (EliteOCRReader as-was)

  1. #1

    RegulatedNoise - trade tool with zxctypo's best-in-class OCR (EliteOCRReader as-was)

    RegulatedNoise is the combination of EliteOCRReader and zxctypo's EliteBrainerous OCR engine. It can OCR your prices, analyse them, show them on your phone/tablet via a webserver, and a bunch of other stuff.

    User Manual (work-in-progress): https://github.com/stringandstickyta...20User%20Guide.

    You can just unzip new releases over the top of old ones - the four config/settings/user data files are valid between versions. Binaries and source code at GitHub: https://github.com/stringandstickyta...Noise/releases.

    v1.8
    * Add MRmP's editable calibration system
    * Add magnifier for calibration purposes
    * Remove incredibly-dumb bug that has been eating the spaces in station names... d'oh...
    * Check for empty string from EDDN, and at least don't crash on general malformed data

    Please feel free to log issues at GitHub.

    Q: Why is your Git archive full of binaries?
    A: Because time

    Q: Windows SmartScreen?
    A: You may need to right-click the file, go to Properties and Unblock it

    Q: Crash when grabbing a screenshot?
    A: OCR functions REQUIRE the Visual Studio 2012 C++ Redistributables for BOTH x86 and x64, from here => http://www.microsoft.com/en-us/downl....aspx?id=30679

    Q: Crash when starting the EDDN listener?
    A: The ZeroMQ component I use for EDDN listening, require the Microsoft Visual Studio 2013 C++ Runtimes. You can get them here => http://www.microsoft.com/en-gb/downl....aspx?id=40784

  2. #2
    I get this error when I try to use this:



    I have the C++ Redis installed for both x86 and x64.

  3. #3
    Originally Posted by Highrisedrifter View Post (Source)
    ArgumentOutOfRangeException ... Ocr.cs line 154
    Hm. It hasn't managed to find any rows to load. Can you post or link to the screenshot you were using, please? Also your calibration.txt values.

  4. #4
    I got this too, but it was fine after recalibrating ocr

  5. #5
    One question: How can I make the tool learn?
    Every import it gets fruits wrong among other words.
    I add the correct name but it continue to happen.

  6. #6
    Originally Posted by Lyker View Post (Source)
    One question: How can I make the tool learn?
    Every import it gets fruits wrong among other words.
    I add the correct name but it continue to happen.
    Can you give me some exact examples please? Ideally post a link to a screenshot. I can easily tweak the algorithm that fixes commodity name OCR errors, but I need specific examples.

  7. #7
    Working really nicely for me, nice work. Only thing is the Autosave is always empty of any data and so I need to restart from scratch every time.

  8. #8
    Originally Posted by Smacker View Post (Source)
    Working really nicely for me, nice work. Only thing is the Autosave is always empty of any data and so I need to restart from scratch every time.
    That's odd. Are you closing the app in an unusual way, with Task Manager for instance? I'll add extra logging code to the autosave in the next version - in the meantime you can use "Export unified CSV" to save your data manually, and "Open CSV" to reload that file next time you start the app...

  9. #9
    Originally Posted by maxh2003 View Post (Source)
    That's odd. Are you closing the app in an unusual way, with Task Manager for instance? I'll add extra logging code to the autosave in the next version - in the meantime you can use "Export unified CSV" to save your data manually, and "Open CSV" to reload that file next time you start the app...
    I think it is to do with the include extended info in CSV flag, I had that unchecked. Now it is checked it seems OK.

  10. #10
    Originally Posted by Highrisedrifter View Post (Source)
    I get this error when I try to use this:

    see #2
    Same 4 me....

    Btw, could I "train" the OCR to read German screenshots (e.g. "Silber" = "Silver" or "Kaffee" = "Coffee") or would I have to switch the ingame language to English if I want to contribute? If not I would just have to take the EDDN data provided by others...

    Keep up the good work!
    Maverick

  11. #11
    Originally Posted by MavMcLeod View Post (Source)
    Same 4 me....
    Well, I could probably fix it but I need you to post an example screenshot. Maybe post it to imgur.com and link it here? If no-one posts a screenshot that shows the problem, I can't fix it! Ideally I need your calibration.txt values too

    Btw, could I "train" the OCR to read German screenshots (e.g. "Silber" = "Silver" or "Kaffee" = "Coffee") or would I have to switch the ingame language to English if I want to contribute? If not I would just have to take the EDDN data provided by others...
    That's a good question, and probably that doesn't have an answer yet, EDDN still being in its infancy. I can imagine adding functionality so that you (the user) can create a translation dictionary - that way your data could be auto-translated as it was uploaded to EDDN - but is it worth the effort of me creating and users maintaining such a dictionary system? I'm not sure :-/

  12. #12
    Originally Posted by Smacker View Post (Source)
    I think it is to do with the include extended info in CSV flag, I had that unchecked. Now it is checked it seems OK.
    Brilliantly-spotted, you are absolutely correct. This will be fixed in the next version

  13. #13
    You got PM max.

  14. #14
    Just a stupid question is there a chance of getting a colum that show us the distance ( farest and closest station )? Or is there one and I haven't noticed it ?
    Thanks for the answer !

  15. #15
    OOoo I just stumbled across the new version! Very nice! Will check this out in a bit Yey!

    Hope that the OCR is as reliable as the EliteOCR tool I've been using until now. Like that this all becomes so much more automatic which is great! Quick question, is there an option so that after it has done the OCR scrape on the screenshots that it deletes them afterwards so that I dont need to manually go in and purge the screenshot folder several times a day?

    Maybe once the whole import is finished you click a save button or something and it asks if you want to purge the screenshots used for the OCR batch?
    Great work btw.

    - - - - - Additional Content Posted / Auto Merge - - - - -

    Another suggestion which I'm finding difficult to do at the moment using the OLD tool.
    I am trying to work out routes which provide the highest RETURN trip profit...

    Using the current tabs I am able to work out for example that 4 of the station I am using will provide a one way trip of 1300 profit per tonne. However I am more interested in finding the highest to and from return trip.

    I.e. a return trip tab which will allow me to sort the overall trip profit so I can see which stations provide the highest TO and FROM profits. Not sure if that makes sense?

    When it does the OCR I get a confirmation window where I correct the misdetected words with a number next to it. Does the app learn every time I make the correct change here for the next OCR run through or will I need to manually enter every time here.

Page 1 of 32 123411 ... LastLast