Computer Vision for Football Analysis in Python with Yolov8 & OpenCV

Поделиться
HTML-код
  • Опубликовано: 21 ноя 2024

Комментарии • 61

  • @ahmedboulahia6033
    @ahmedboulahia6033 11 месяцев назад +8

    Great content! The result is so professional 👏 👏

  • @marcounipd1836
    @marcounipd1836 3 месяца назад +3

    Searching for a video like this from months, great content!

  • @TauquirAhmed-pe2ou
    @TauquirAhmed-pe2ou 7 месяцев назад +1

    Amazing project with great explanation

  • @MichaelMacDonald-im4bq
    @MichaelMacDonald-im4bq 5 месяцев назад +7

    Kognia Sports Intelligence has built this 5 years ago, and has 4 global patents. In fact, not only does Kognia detect "events" and "physical" stuff, but also can detect 100s of tactical soccer movements (example: when players have movement behind the defensive line, etc.) it's very advanced!

  • @hamidGghassrimanesh
    @hamidGghassrimanesh 10 месяцев назад +2

    Amazing! Professional.❤

  • @fadytarek4711
    @fadytarek4711 9 дней назад

    Thank you for this video! Very interesting to me, if I have my own Input video how can I train the model to detect the players?

  • @bayda3695
    @bayda3695 8 месяцев назад +1

    great work. can you upload the dataset to the Roboflow so we can download it easily?

  • @beattiewarren
    @beattiewarren 8 месяцев назад

    Brilliant! Respect!

  • @vivekgangurde9685
    @vivekgangurde9685 8 месяцев назад

    Great work u have provided me great insight for my project its really helpful thank you 😊

  • @troysm2146
    @troysm2146 3 месяца назад +1

    where can i get other videos to use instead of the one you have linked is there any websites also do i have to use a certain fixed camera for the input video
    thanks in advance

  • @patrickmcc
    @patrickmcc 3 месяца назад

    Really insightful. Would this be possible with live streamed video of a match? If you were using the same fixed camera?

  • @raylenders1858
    @raylenders1858 6 месяцев назад

    Hey, good work reducing computing data when you reduced the homogenous matrix update frequency.

  • @bayda3695
    @bayda3695 8 месяцев назад +1

    can we use google colab or host the web application on a server with GPU to improve the tracking speed?

  • @abrahamceviz
    @abrahamceviz 6 месяцев назад

    Türksünüz herhalde. Aksanınız hoşuma gitti :) Dinlemek oldukça güzel sizi

  • @minuanand6916
    @minuanand6916 7 месяцев назад +2

    hello , I am stuck in an project where I have to track the movement of a cricket ball , so can you help me in that?

    • @samridhgoel752
      @samridhgoel752 6 месяцев назад

      i also wanna build a something similar luke this in cricket, maybe we can help each other

  • @obsidian_fury33
    @obsidian_fury33 8 месяцев назад

    Great work my man, thnx for the share.

  • @Kishi1969
    @Kishi1969 6 месяцев назад

    Dear Sir,can the project also be using for other detection, like these 3 parameters ( Bus,car and truck) with datasets annotated already and video are available,if possible let me hear from you Sir, once again you have done a great job

  • @abdelmajidkansoussi6904
    @abdelmajidkansoussi6904 7 месяцев назад

    Great work Mr Hamza. Please upload more content

  • @punkysuen
    @punkysuen 5 месяцев назад

    Is it possible to assign field keypoints on 180 degree camera video to have a proper detection? The field does not show straight line on 180 degree video.

  • @mdriad4521
    @mdriad4521 7 месяцев назад

    great man! we want more end to end project like this.

  • @graphguy
    @graphguy 3 месяца назад

    Is this the same code for TENNIS?

  • @nguyen_van_thang_28
    @nguyen_van_thang_28 5 месяцев назад

    thank you so much. Where can I get the dataset, sir?

  • @kappaebi6802
    @kappaebi6802 8 месяцев назад

    Thank you for this video! Very interesting to me. Would you tell me how you got technical recording game? I want to get recent technical recording games in Premier League.

    • @MLWH
      @MLWH  8 месяцев назад +1

      I only found a few recordings on youtube. I don't know where you can find more recent ones.

    • @kappaebi6802
      @kappaebi6802 8 месяцев назад

      Thank you so much! I try to look for them.

  • @zorananastasius
    @zorananastasius 7 месяцев назад

    what website was used for the Footbal Players Detection ??

  • @BerkerTomac
    @BerkerTomac 8 месяцев назад

    Thank you for this great content! Is it possible for you to share the datasets for Players and Keypoints, couldn't see them in your repository. Once again brilliant video.

    • @MLWH
      @MLWH  8 месяцев назад

      I can't add the dataset to the Github repo because of its size. You can email me and I'll send you the dataset.

    • @exitiumsn9064
      @exitiumsn9064 8 месяцев назад

      @@MLWH hey could i get the datasets aswel please

    • @wojciechwoszczek1484
      @wojciechwoszczek1484 8 месяцев назад

      @@MLWH Hey, I've also send you an email with request of sharing the dataset :) I'm looking forward to hearing from you, your video inspired me!

    • @user-zw7zd3qv1w
      @user-zw7zd3qv1w 7 месяцев назад

      ​@@wojciechwoszczek1484Did you receive the dataset?

    • @user-zw7zd3qv1w
      @user-zw7zd3qv1w 7 месяцев назад

      ​@@exitiumsn9064Did you receive the dataset?

  • @alirezaghasrimanesh2431
    @alirezaghasrimanesh2431 9 месяцев назад

    amazing!

  • @BALONESDEOROS
    @BALONESDEOROS 3 месяца назад

    hey bro, thanks for your video,. do you know an app to create and see the skills or goals from diferent angles?

  • @ihamzasadiq
    @ihamzasadiq 4 месяца назад

    How long can the video be?

  • @tıbhendese
    @tıbhendese 6 месяцев назад

    I want you to make a different "DL in medicine" project than other channels. People make them in short, lack of deeply understandable, lack of math, lack of information, and inefficient ways. Most of them don't share datasets, or make it very short&fast. Some type the code fastly and just run it in beginner level.
    Make an at least 2 hours of project video which includes information of problem, models used, library functions and coding step by step. And please share the dataset in the end. Medical classification or clustering. Show us the image processing and model tuning

    • @yagzsoylu8653
      @yagzsoylu8653 6 месяцев назад

      blackseastormu nerden biliyorsun merak ettim

    • @tıbhendese
      @tıbhendese 6 месяцев назад

      @@yagzsoylu8653 internette aratınca şarkıcı çıkıyor ben onu bilmiyordum xD sadece karadeniz fırtınası yazmak istemiştim eski ve hızlı zamanlardı :P kaldı öyle

  • @pkmnjourney
    @pkmnjourney 10 месяцев назад

    what if we wanted to also track the player?

    • @MLWH
      @MLWH  10 месяцев назад

      To do that you can either use the tracking capabilities of Yolov8, or implement a tracking algorithm like deepsort in the loop

  • @alexanderwijayanto2780
    @alexanderwijayanto2780 6 месяцев назад

    Mantap bro 😊

  • @altafurrahman9404
    @altafurrahman9404 9 месяцев назад

    You deserve a subscribe atleast

  • @abuyusufmushtaha9014
    @abuyusufmushtaha9014 6 месяцев назад +1

    يا حمزة نزل بلعربي الاجانب كثار بالمجال عكس العرب

  • @abdellatif.x8127
    @abdellatif.x8127 6 месяцев назад

    Your are Moroccan ?

  • @nadjibbensaidane6552
    @nadjibbensaidane6552 6 месяцев назад

    حمزة معندكش لي فيديو بل عربية ؟

    • @TahaOummani
      @TahaOummani 6 месяцев назад

      فين اخاي مزيان
      ممكن تعطيني فيسبوك ديالك ندوي معاك الله يحفظك ❤

  • @thebudaxcorporate9763
    @thebudaxcorporate9763 7 месяцев назад

    streamlit power

  • @arupde6320
    @arupde6320 7 месяцев назад

    be regular

  • @thomasbradley2916
    @thomasbradley2916 11 месяцев назад

    why yolov5 and not yolov8?

    • @MLWH
      @MLWH  8 месяцев назад

      🤔 It's Yolov8

  • @thiagopereiradossantos588
    @thiagopereiradossantos588 6 месяцев назад

    I think it's very weak , I can do better

  • @StupidusMaximusTheFirst
    @StupidusMaximusTheFirst Месяц назад

    It makes a few mistakes here and there, mistaking one team for the other in some players. It'd be funnier if it made mistakes like mistaking them for a team not even playing, or if it detected players as refs, or more than one ref etc. LOL 🤣

  • @woofy4852
    @woofy4852 8 месяцев назад +1

    Hi bro, do you have any discord? cuz i need help with this code T-T

    • @MLWH
      @MLWH  8 месяцев назад

      I don't use discord. But you can email me your questions, and I'll try to clarify the code for you.

  • @shekharmeena529
    @shekharmeena529 9 месяцев назад

    How can we detection actions like foul, yellow card, red card, goal, corner in the real live match using this

    • @MLWH
      @MLWH  8 месяцев назад

      For that you need to add a model for football event detection.

    • @jayeshbaviskar9251
      @jayeshbaviskar9251 8 месяцев назад

      @@MLWH can you give a overview of how it can be implemented

    • @MLWH
      @MLWH  8 месяцев назад +1

      @@jayeshbaviskar9251 The simplest approach is to train a model that takes as input a sequence of frames and predicts the event happening.