CHESS MOVES DETECTION USING COMPUTER VISION ALGORITHM

Поділитися
Вставка
  • Опубліковано 22 тра 2024
  • Many people have addressed this problem using various different ways, and here is mine.
    My algorithm uses techniques to computer vision and machine learning to detect the chess move on the live chess board.
    The video is mainly to describe my approach and request all of yours valuable suggestions on this problem. It addressed various aspects around this like why, how and what?
    This is not perfect yet, there is a long way to go....
    I want to thank everyone involved in this. Specially my father and mother to teach me chess at very young age and support me in all situations, my wife Rutu who was behind the camera for this video and her immense motivation helped me a lot throughout this journey.
    VIDEO CHAPTERS
    0:00 Demo
    0:51 Disclaimer
    1:46 Motivation (Why?)
    4:05 Approach (How?)
    5:27 Camera Position
    7:05 Detecting the corners
    8:50 Algorithm
    14:28 Machine Learning Model
    18:19 Text to speech
    19:34 Conclusion
  • Наука та технологія

КОМЕНТАРІ • 39

  • @JayshreeParmar-ke8hu
    @JayshreeParmar-ke8hu 10 днів тому +1

    Super & great work 🎊🎉

  • @aditya.21
    @aditya.21 7 днів тому +2

    Really loved it! Fits the domain space perfectly for me pursuing Computer Science and playing Chess since a long time now! Now I feel I must give it a try as well for my project work😀😀

  • @jaysolanki9766
    @jaysolanki9766 10 днів тому +1

    Amazing 🤩. Great work Sachin Bhai .

  • @zeelparmar2301
    @zeelparmar2301 17 днів тому +3

    Absolutely love seeing technology push the boundaries of traditional games like chess. Python's role here is not just impressive but inspiring. Amazing work! 👏

  • @DEEPAKKUMAR-ch6cb
    @DEEPAKKUMAR-ch6cb 10 днів тому +1

    Very Great Bhaiya. I absolutely respect your efforts you have put behind this..

  • @nirurathod266
    @nirurathod266 17 днів тому +2

    Very good innovation

  • @rutuparmar60
    @rutuparmar60 17 днів тому +3

    I think your innovation is very useful in future 😊bright future of chess by computer vision

  • @Yogesh.Janxatriya
    @Yogesh.Janxatriya 17 днів тому +2

    સચિને ખૂબ સરસ અલ્ગોરિધમ બનાવ્યુ છે.
    very Big congratulation....we proud of you. exellant work..

  • @kaushalvyas4934
    @kaushalvyas4934 17 днів тому +2

    Good Innovation 👏🎉🎉

  • @bhavnaparmar9700
    @bhavnaparmar9700 17 днів тому +2

    Excellent

  • @mishhvaa_2625
    @mishhvaa_2625 16 днів тому +2

    Wowwwwww 🎉

  • @nimishapanchal897
    @nimishapanchal897 16 днів тому +2

    Very good bhai

  • @nimishapanchal897
    @nimishapanchal897 16 днів тому +2

    Excellent innovation

  • @RipalChauhanB
    @RipalChauhanB 17 днів тому +2

    Excellent , innovation 🎉

  • @sudhamjayanthi
    @sudhamjayanthi 12 днів тому +2

    was just thinking of building something similar recently and i see this video popup on my feed, great stuff!

  • @umangchaudhary7522
    @umangchaudhary7522 17 днів тому +2

    Wonderful work 👏

  • @SadhuPrasanga
    @SadhuPrasanga 16 днів тому +2

    Good work, but can you try implementing it with any camera position and distance? It should not be difficult. Moreover, if the game is tracked by the camera from the start, it should not be difficult. Also, what if the camera misses a move or two by some chance. Can the code report the lapse? Also, can it report an illegal move? Can it report a draw based on 50 move rule and three fold repetition rule and thrice same position rule? Can you connect the code to a chess database to find if it's a new position not played before?
    You should be interviewed by Chessbase India. You can contact them if they don't find you.

    • @sachinschesschannel2627
      @sachinschesschannel2627  16 днів тому +3

      First of all, thanks a lot for your detailed observations!
      For Chess related things like "report a 50 move / 3 fold draw", "connecting with database" -- Yes, these all are defiantly doable and actually straightforward!
      For the other things related to Camera position, Move miss, illegal move: This all are challenging and perhaps a bit subjective to implement. And performance depends on lot of external factors. I need to put more thoughts into this, will get back to you!
      I will try to reach out to Chessbase India, I am their long time follower and a big fan by the way!!

  • @AnshBajpayee
    @AnshBajpayee 12 днів тому +2

    Hey bhaiya,
    I'm fresher,very keen to learn openCV. From where you've learnt to buikd theese models and implement it using open CV . Can you please tell me about any lecture present on youtube or a course to teach me how building these models works.
    Thank you.

    • @sachinschesschannel2627
      @sachinschesschannel2627  12 днів тому +2

      Hi, Thank you for your comment!
      There are tons of resources to learn openCV. But it depends on what you want to solve using it exactly.
      I have not followed anything particular for this except for the given doc strings in the code and the official documentation: pypi.org/project/opencv-python/
      If you have decent coding skills and you know python than your can start with this: www.geeksforgeeks.org/opencv-python-tutorial/
      There might be better resources which I might not be aware of, but it really doesn't matter from where you learn.

    • @AnshBajpayee
      @AnshBajpayee 12 днів тому +2

      @@sachinschesschannel2627 Thank you bhaiya, I code in c++ and do cp and have a rating of 1450 on codeforces.
      But I wanted to explore this domain so asked your advice. Thank you for this.

  • @akashgupta-pb2kd
    @akashgupta-pb2kd 15 днів тому +2

    Yo bro

  • @shubhsardana8040
    @shubhsardana8040 16 днів тому +2

    I have some questions. how to reach out to you?

    • @sachinschesschannel2627
      @sachinschesschannel2627  15 днів тому +1

      Sure, Thanks!
      Email: sachinchauhan10497@gmail.com
      LinkedIn: www.linkedin.com/in/sachin-chauhan-chess/
      Note: Response time might be around few days!

  • @pulkitsujaan
    @pulkitsujaan 15 днів тому +2

    Hey, Can you please share the source code?