Latest Pose Estimation Realtime (24 FPS) using CPU | Computer Vision | OpenCV Python

Подписаться 409 тыс.

Просмотров 176 тыс.

50% 1

In this tutorial, we will learn Pose Estimation. We will detect 33 different landmarks within a human body and all of this will be done in real-time. That's right more than 24 fps. We will first look into the basic code required to run and then create a module out of this so that we don't have to write the code again and again for different projects.
🚀🚀 My Urdu/Hindi AI RU-vid Channel 🚀🚀
/ @murtazahassan01
Download Code:
www.computervi...
Premium Courses:
✔️ Computer Vision Game Development Course:
bit.ly/3ttLZ2s
✔️ Computer Vision with Arduino Course:
bit.ly/3wzLB4m
✔️ Advanced Drone Programming Course:
bit.ly/3qs3v5g
✔️ Learn to Build Computer Vision Mobile Apps:
bit.ly/3uioY1J
✔️ Jetson Nano Premium Course:
bit.ly/3L8uIlF
⚙️⚙️⚙️--My Gear - ⚙️⚙️⚙️
👉 Complete Gear 💈: www.computervi...
👉 My PC Specs 🖥️: www.computervi...
👉 My Video Shooting Gear📽️ : www.computervi...
👉 My Laptops 💻: www.computervi...
👉 Educational Products🧑‍🎓: www.computervi...
👉 RU-vid Starter Kit 🔴: www.computervi...
Follow Me:
TikTok: bit.ly/3Vo76OQ
Facebook Group: bit.ly/3irDcb7
Discord: bit.ly/3JvyxAM
Facebook Page: bit.ly/3IvpU7W
Instagram : bit.ly/3NdGME3
Website: bit.ly/3ICFTS0
Github: bit.ly/3woU6PS
#ComputerVision
#OpenCV
#CVZone

Опубликовано:

11 сен 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 226

@andrespereira4852 3 года назад

Murtaza, OMG I am almost speechles with every video posted of yours. You are bringing HUGE aid steps for every developer of tomorrow. Thanks to you and have a great day!!!

@harshmirdhwal 3 года назад

0:20 a module out of it!! I've listened it like 10 times, is your intro a rap or song, Awsm man! please keep making these kinds of video and never stop

@Shizuko-Zen 3 года назад

It's a crime that your channel has less than 1M subs. Absolutely amazing stuff! As a CS student, content like this really helps me put what I have learned into some practice (or al least gain some perspectives).

@fabulousfox9057 3 года назад

This vid makes motion capture without spending money finally possible!

@TheSpurious 3 года назад

Hey, thanks for not shoving everything into a jupyter notebook. Using real files has always been the best way to learn for me.

@chisomokwor8133 3 года назад

Thank you for making this amazing video! You're a lifesaver. I'm doing a Computer Vision senior project, and your channel has literally been the most helpful! Keep being amazing

@murtazasworkshop 3 года назад

Glad it helped!

@jimmy00080002 2 года назад

Hi I would like to ask for beginners, what is the learning path for this?

@laus-thecurious4120 2 года назад

@@murtazasworkshop your code is not visible on the site . it needs enrollment . but you didnt add any enrollment button on the webpage .

@jsims4028 3 года назад

Excellent video as always. Sir Murtaza is my hero and a world reference in OpenCV, he deserves a Nobel Prize

@XX-vu5jo 3 года назад

Lol he’s just using others work in his videos LOL

@antonykathage7199 3 года назад

This is honestly something I've been trying to work on for a while now. And you've just shown me how! I found your channel on the hand tracking video and absolutely loved it and integrated it with an existing project of mine. Thank you so much!!

@PunxTV123 3 года назад

the best channel in programming right now, i mean yes cause he's using latest tech as guide and I learn a lot

@mohammedyaheyawaleed7445 3 года назад

hi first of all I want to thank you murtaza for your efforts and the thing I want to tell is this In the def __init__(self):------(method) upBody=False is giving an error so I checked and now we have to use model=1 and in place of self.upBody = upBody we have to write self.model = model Here's the new code: class poseDetector(): def __init__(self, mode=False, model=1, smooth=True, detectionCon=0.5, trackCon=0.5): self.mode = mode self.model = model self.smooth = smooth self.detectionCon = detectionCon self.trackCon = trackCon self.mpDraw = mp.solutions.drawing_utils self.mpPose = mp.solutions.pose self.pose = self.mpPose.Pose(self.mode, self.model, self.smooth, self.detectionCon, self.trackCon) it took me 30mins to solve this because I am a beginner. Others may also get this error So, I thought to share it with you

@npraam5176 2 года назад

still im getting the same error brother

@Batselot 2 года назад

@@npraam5176 I actually found the cause of it. When you check mediapipe's integration of the code, they aren't using upBody or any other argument like that. Remove detection con and others and their implementations from the code as well and the code should work just fine.

@saivamshikrishnasarala2268 2 года назад

@@Batselot bro can u share me the code here please ?

@matthewdavis182 Год назад

Change the init to this: def __init__(self, mode=False): self.mode = mode self.mpDraw = mp.solutions.drawing_utils self.mpPose = mp.solutions.pose self.pose = self.mpPose.Pose(self.mode)

@piyushgupta9072 Год назад

this is also giving error try this one : def __init__(self, mode=False): self.mode = mode self.mpDraw = mp.solutions.drawing_utils self.mpPose = mp.solutions.pose self.pose = self.mpPose.Pose(self.mode)

@edSabio572 3 года назад

Murtaza, Thanks for making a complex subject so easy to understand, a mark of a great teacher!

@michellebaltazar2447 3 года назад

Hello I just want to say that you've been very helpful in my projects and research work. Coming from a little programming background, your tutorials helps a lot. Thank you so much!

@nithinsai2250 3 года назад

do u know where to get pose module?

@nuca5104 2 года назад

Thank you so much! You don't realise how much you've helped everyone on here... nothing like this on youtube.

@chandank5266 3 года назад

That's what i was waiting for.. thanks Murtaz

@grgkaralis 2 года назад

Great content. In clear steps. Writing the code following pep8, explaining each function of the module in use, refactoring afterwards, providing accurate info about oop. Flawless

@engenglish610 3 года назад

Upvoted for this work. However, I have a question : can you detect two persons in the same video ?

@vaibhavrajnathchauhan5340 3 года назад

Sir your videos are gems, must needed tutorials in lockdown

@manishdhal5073 3 года назад

You're a Legend living among mortals! Keep the good content up!!!

@geraldmusandirire5446 3 года назад

Best pose estimation explanation and implementation video on the internet. Had been trying to understand this for some time and you explained it well

@nithinsai2250 3 года назад

do u know where to download that poseModule?

@loicsan5616 3 года назад

@@nithinsai2250 have you the answer to your question? please me too i want the folder

@tazfranck 3 года назад

Thanks Murtaza ! Your video is very clear and so useful ! Keep Going on !

@dsebooksellernotes1529 2 года назад

TypeError: create_bool(): incompatible function arguments. The following argument types are supported: 1. (arg0: bool) -> mediapipe.python._framework_bindings.packet.Packet When I run the code, it showed the above error. Can anyone help me fix the problem? Thanks

@sayasaya.mp4 2 года назад

Hiiii👋🏻 Could you find the solution? I'm facing the same issue right now. Can you help me please?

@matthewdavis182 Год назад

Change the init to this: def __init__(self, mode=False): self.mode = mode self.mpDraw = mp.solutions.drawing_utils self.mpPose = mp.solutions.pose self.pose = self.mpPose.Pose(self.mode)

@nnamdiekweariri 3 года назад

Hi, thanks for your wonderful videos and projects. I find this particular video very useful since it's related to what I'm currently building. One problem with media pipe though is that its unable to detect multiple persons at once. For this I've used OpenPose, which is much more complicated.

@mayanksingh1345 3 года назад

Thank you God. I found your channel. Thank God. I want to learn opencv and them work for Tesla when they come in India. Sir please tell me do you teach on Udemy also?

@davidallan6036 3 года назад

Explained very well, I look forward to each of your videos

@Piotful Год назад

Hi Sir, So good work. Could you tell me where to find the sample videos used in the course cv2.VideoCapture('PoseVideos/3.mp4'). Thx

@phongthe1272 3 года назад

Many thanks for the Good tutorial for learning something like: Pose tracking, Module's structure and Python code.

@harshmirdhwal 3 года назад

Man You're rockin' opencv

@Kishi1969 3 года назад

Good lecturer and keep it up for people like but kindly include the video you used to demonstration in your program, so that we can test it with it as well

@donteo3105 3 года назад

you had great teaching materials thanks... hope you can a video on Computer Vision trigger action movement for Robotics Arm autonomously... thanks

@siddharthasiddhartha5471 2 года назад

hi murtaza, i really love your videos.. very informative... kudos man

@yehe501 3 года назад

fantastic video, well done good job

@bas_abhi 3 года назад

This is so cool!! Really helpful and amazing!!

@belkassem06 3 года назад

The Genius struck once again. Thanks a lot.. @Chan Xu, I AGREE 10000%

@arjunpalusa9421 Год назад

Just loved it, thanks a lot, God bless you.

@cristobalfigueroa8413 3 года назад

Best channel!

@creativecore3575 2 года назад

found it ! Thank you, Sir. Amazing content as always !!

@Bartek9391 3 года назад

Amazing and helpfull! Thank you!

@ghaithmechi6858 2 года назад

Hi, thanks for your wonderful videos and projects.

@__________________________6910 3 года назад

100k sub soon

@bw3047 2 года назад

Probably the best app from you..

@ThongNguyen-gy4ng 3 года назад

its really a great video, but its only detect 1 person. How can i detect many person!

@carlosoliveira1922 3 года назад

Great video!!! Thank you so much for all the knowledge provided by each video that you produce! May God always keep you health and prosper!! I have one questio if you permit me, what would happens with a video with 2 people?

@nshanthi7668 2 года назад

Thank you so much for your amazing videos. Please upload a video on python OOPS concept.

@gustavomonteiro4814 3 года назад

THANK YOU SIR! That helped A LOT!

@rohitsonawane4592 3 года назад

Can we use hand detection and pose detection at the same time.

@laus-thecurious4120 2 года назад

your code is not visible on the site . it needs enrollment . but you didnt add any enrollment button on the webpage .

@rog0079 3 года назад

great video! what if we had multiple humans? what settings we will need to do in order to detect the poses correctly of each human?

@luisduarte6456 2 года назад

Congratulation, excelent video very useful.

@sharathyanamandra 4 месяца назад

Mediapipe works for a single person. what if i want to build a multi object detection (in our case multi human detection and multi pose estimation model). could you explain how to build the model architecture from scratch ?

@dhias5314 3 года назад

Thanks , Your video is very useful, does this use openpose?

@pttangent211 3 года назад

thank you so much, this is a great video that will help me in graduation projects, could I get the reference article of the pose estimation model used by this video? I should learn more about the details of the mechanism thank you again for your great work!

@user-cl7fv5gi3o 3 года назад

You're the best❤️❤️❤️❤️❤️👍🏻👍🏻👍🏻

@high-jump-marvel 3 года назад

Hi, where is video? cap = cv2.VideoCapture('PoseVideos/1.mp4') How to get 1.mp4? Thnaks for reply...Isaco

@mateuscardoso7626 9 месяцев назад

ERROR: Could not find a version that satisfies the requirement mediapipe==0.8.9 (from versions: none) ERROR: No matching distribution found for mediapipe==0.8.9

@madebymate4870 3 года назад

Thanks dude! Really cool tutorial Helps a lot understanding opencv Altough I still have no idea how can you do this while screen recording So if you or anyone knows I hope you can help Thanks!

@rajmeetsingh1625 Год назад

Hi Sir, please tell me the difference between openpose and mediapipe library. Which one is better ?

@leonardosouzaconradodesant6213 Год назад

THANK YOU FOR SHARE!!!!!

@MM-fv1pi 3 года назад

Which method does this package utilize to discover body parts and then estimate the body posture?! Seems to be working quite rapidly and accurately. I am talking about the basic algorithm put to use by Python CV2 module.

@grapesalt 3 года назад

Hey, I had a question. It would mean a lot if you or anybody else could help. how can we add an image over each body part if we store the data from each frame?

@rondamon4408 3 года назад

Great video, I don't know how you have time to make them

@jcmschott1895 3 года назад

Your name is awesome, dude! Cheers to 'Seu Madruga'!

@__________________________6910 3 года назад

Thanks, make more project on mediapipe

@msk.universe 3 года назад

OMG this is tremendous!!!!

@marii0oo_679 2 года назад

HI~Thx for ur wonderful videos. And there are some question i wonder to know is how can i make this pose Estimation on my camera with real time image?

@PiononoMatic 8 месяцев назад

this is awesome! how can i export te data to Auto Rig Pro in Blender?

@AtanuShome 2 года назад

How are you going to the source of POSE function? My VS isn't linking the method with its source while writing the code.

@ahmadaladawi1633 2 года назад

Great, thank you! could you share the link where we can get the videos from?

@davidf8912 2 года назад

Hello, I got your code from the Computer Vision link. However when I run it I get this error "File "C:\Users\dfern\anaconda3\envs\sket1\lib\urllib equest.py", line 1421, in unknown_open raise URLError('unknown url type: %s' % type) urllib.error.URLError: " Any suggestions?

@yardendror9023 3 года назад

Great Video, I would be glad if you could send me the how to solve the error which appear when the videos ends.... thans alot!

@MuhammadUsman-wh4zi 3 года назад

Thank You sooooo much Brother..

@rishabhsharma7261 3 года назад

Amazing!!

@dhruv.pandey93 3 года назад

I already have a Raspberry pi 4B 4Gb with me. Should I instead switch to Nvidia Jetson Nano 2gb? I know Raspberry Pi does not have a good GPU as compared to Jetson. But do we really need a GPU for open cv based computer vision(example: the volume control) or is it only needed for training machine learning models? Is it ok that I train my models on Kaggle or Google Colab and then deploy it to Raspberry Pi. Would I still need the GPU even if I am not training on Raspberry and only using it for detection or classification?

@doguhanaslan7333 4 месяца назад

Where did we create the pose() function, I think I missed it

@actionkey8042 3 года назад

Murtaza's Workshop - Robotics and AI awEsoOOOOme and fuUUUnny

@matthewdavis182 Год назад

If anyone is getting an error about: 1. (arg0: bool) -> mediapipe.python._framework_bindings.packet.Packet Change the init to the following code: def __init__(self, mode=False): self.mode = mode self.mpDraw = mp.solutions.drawing_utils self.mpPose = mp.solutions.pose self.pose = self.mpPose.Pose(self.mode)

@TheBlackenedStar 3 года назад

Is there a way to download the example videos that you use for the demonstration?

@krishanchandra443 3 года назад

Thank you sir for the tutorial. I want to ask something how do we save the video using opencv. Thank you

@aimanyounis8387 3 года назад

Hey, thank you for this awesome work. Can I use skeleton to count steps for person?

@wgalloPT 3 года назад

I have a code editor and learned some javascript in the past. I know nothing about Python. But im in healthcare and would really like to run this. If i follow the instructions above do you think I can do this? do you think i can make this work if I know nothing about Python? I enrolled in the course. I really want to get this to work for me. What do you suggest I need?

@rajmeetsingh1625 2 года назад

Sir, I need an angle between legs during walking. Please share some work related to finding angle estimation. I am waiting.

@atharvakatre3256 3 года назад

I am not able to run the findAngle() method from your code, getting list index out of range error. Can you please help with this

@user-cw8hd7xn1i 2 года назад

Hello, Thank you for your project. I planned to make program that analysis human pose in RealTime!! But, I have some problem,,, Where is the mp4 file? I can't find video files.

@ronyra 3 года назад

this is amazing! i have a question, how can i add classification to it so it can detect a different fitness exercises like mountain climbers, jumping jacks ect...?

@hichem1123 3 года назад

a CNN with your differents positions and their labels ? you train a model and then you get what you need, or try simply to add a print from word' "name of exercice" in an array simply when the action == to word

@amithadole9252 2 года назад

After detecting pose how we can find its siting, standing or laying? Please help me to solve this.

@inspbushi 3 года назад

Hi, thank you very much for the video! I tried recreating the code but I have an issue when I create it as a module; For some reason the speed of frames decreases by almost 5 times and even it becomes 0 at times... Not sure how to fix this, could you please give some insight?