Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Open CV and Java

Options
  • 20-08-2019 9:40pm
    #1
    L.Jenkins
    Moderators, Computer Games Moderators, Technology & Internet Moderators Posts: 19,240 Mod ✭✭✭✭


    Anyone have any experience using Open CV, with or without Java? I suppose this might tie into machine learning somewhat, as I'd like to dabble in object recognition. Not pulling text from an image or PDF, but actual object identification and recognition.

    I'm intent on working on an open source project, where pictures of litter are taken and uploaded via mobile app or to a website. Most of the infrastructure is already there, but I'd like to help and automate the process of verifying waste by category and brand.

    I have browsed the Open CV website, but I'm also looking for additional tutorials or resources if anyone knows of good resources.


Comments

  • Options
    #2
    TuringBot47
    Registered Users Posts: 991 ✭✭✭TuringBot47


    I've used Python to call OpenCV on Windows & Linux.
    You need Python, it's better to support the native nature of OpenCV.

    There's a good website : https://www.pyimagesearch.com/
    He provides snippets of code, but he's training books/vids are pricey enough.

    I've messed around with line detection, image segmentation, processing video including detecting cars in dashcam vids using Yolo.
    https://www.pyimagesearch.com/2018/11/12/yolo-object-detection-with-opencv/

    It's actually surprisingly good at detecting cars/trucks in a dashcam video.
    But depending on the video resolution you'll get anywhere from 1-10 fps processing, not realtime speeds.


  • Options
    #3
    TuringBot47
    Registered Users Posts: 991 ✭✭✭TuringBot47


    Classifying "litter" by category and/or brand would be very tricky I imagine.
    By it's very nature, it will be dirty, probably partially obscured in grass or the like, with no real defined shape.


  • Options
    #4
    smacl
    Moderators, Society & Culture Moderators Posts: 15,750 Mod ✭✭✭✭smacl


    I did Sattya Mallick's Learn OpenCV course earlier this year which I thoroughly enjoyed. Mainly used it through C++ myself, dipping in and out of Python as needs be.

    As above, I'd guess identifying and classifying litter would be difficult as the image recognition is essentially looking for similar objects to known objects from the training data set. You'd probably be best of looking for certain classes of litter such as cans, bottles, crisp packets etc... to start with and then building on those classes to get most commonly found and identifiable litter types.


  • Options
    #5
    L.Jenkins
    Moderators, Computer Games Moderators, Technology & Internet Moderators Posts: 19,240 Mod ✭✭✭✭L.Jenkins


    Thanks guys for the advice. I'll look into the links you've shared and work with Python/C++ instead.


Advertisement