Awesome, well explained. I have this up and running on a Pi4 it should be smoother on my Pi5 when it arrives. More ideas for projects come flooding into my mind. Keep up the good work.
My jetson nano seems to no longer be able to do these things. Need to get my camera working on one of my raspberry pi 5 with a 64mp camera. I now have a 22 node docker cluster hodge podge mix. Haha Keep up the awesome work!
@@kevinmcaleer28 when you show your setup the camera module 3 isn't plugged in but you're using something plugged into USB port? The only reason I say this is because I can't currently find a way to get Raspberry Pi 5 + Camera module 3 to work with OpenCV. Camera module 3 just wont get picked up with VideoCapture(0). Unless you know of a way?
Super interesting video! I played with OpenCV and Tensorflow on a rPi 3. Back then I used a Movidius (now Intel compute stick) to improve the detection speed. The rPi 5 seems to be even faster, but I was curious if you have tried those compute sticks?
Hello Kevin, thanks for the interesting video! Is the hand detector scale invariant? If it is, I am thinking of detecting other objects, not hands. Do I have to train them at first?
i have an arducam 64 mp hawkeye autofocus camera working on bookworm lite 64 bit. Any chance you can make a tutorial for this with those? or do you know if it would work if on bookworm instead of on rasbian OS?
I am running some project with Image processing using raspbberry pi earlier. But I'm a little bit confuse about should I buy a brand new raspberry pi 5 for my project or its okay to just buy a rp 4 coz the budget limit that I have. Is anyone can help me ?
How does this work, Pi 5 with openCV doesn't work with the command cv2.VideoCapture(0), I know this for a fact. Is there a solution to this problem? Because there is nothing in the notes or instructions. Everyone who tries this code on a Pi5 will get an error and a blank video feed. Please help I'm confused.
The same problem happened to me and it's not about Pi 5, the problem is in the new OS (bookworm), OpenCV is working only in the old OS (bullseye) which is not available in Pi 5, i think in OS 32 Legacy only. Please tell me if you found a solution ?
I have the same issue/problem (14 Mar 2024) with the code supplied running on a new Pi5 and the OS Debian GNU/Linux 12 (bookworm) anyone who gets past this please share some info.
OpenCV supports V4L/V4L2. That is the standard for video devices on Linux. “Pi cameras” refuse to support this. That is Raspberry Pi’s problem, not OpenCV’s problem. Found it on the OpenCV site
Oh yes - definitely should work on Pi 5 with Bookworm. Are you using a virtual environment - Bookworm pretty much prefers this way of using Python3 so that maybe why there is an issue?
@@kevinmcaleer28 Sorry for the dumb question, but I am trying to get this to work on a stock/new Raspberry pi 5 with Bookworm I am new to much of this.. Do you have or could you make step-by-step best install/setup guide for a Noob using a RP5 with Bookworm? Im not sure how to use the virtual environment. I would LOVE to be able to do some of the face tracking exactly as you show in your video.. Not 100% how to install any of this correctly. Thank you.
@@kevinmcaleer28 I made environment and made all the installation there but the cvzone is installed normally python I have 3.10 .just brand new pi 5 with the latest fw.
Good information but the thumbnail lead me to believe that you are showing how to train the pi to recognise certain people when a face is detected but alas.
A simple but extraordinary approach to explaining a fairly complicated subject.
Congratulations, your explanation and slides are very complete. Nowadays we can have great materials online for free like yours. Thanks !!1
Awesome, well explained. I have this up and running on a Pi4 it should be smoother on my Pi5 when it arrives. More ideas for projects come flooding into my mind. Keep up
the good work.
Hello! Have you test it on pi 5 already? How it works?) Is it worth it?
@@РомаРома-х7о Unfortunately my Pi5 was delivered broken, I am waiting for a replacement
Could you try to show fps of tensorflow lite and yolo on pi5 ?
My jetson nano seems to no longer be able to do these things. Need to get my camera working on one of my raspberry pi 5 with a 64mp camera.
I now have a 22 node docker cluster hodge podge mix. Haha
Keep up the awesome work!
Hi Kevin. Great video!!!! Could you please confirm what camera you are using with the Raspberry Pi 5. Thanks
The standard camera module 3 from Raspberry pi
@@kevinmcaleer28 when you show your setup the camera module 3 isn't plugged in but you're using something plugged into USB port?
The only reason I say this is because I can't currently find a way to get Raspberry Pi 5 + Camera module 3 to work with OpenCV. Camera module 3 just wont get picked up with VideoCapture(0). Unless you know of a way?
What is the approximate detection frame rate?
Thanks for sharing. You are very talented.
Thanks, I really enjoyed the show. Have now subscribed and will go through old videos.
Br.
Ole Soerensen
Impressive project/speed 🙂
Super interesting video! I played with OpenCV and Tensorflow on a rPi 3. Back then I used a Movidius (now Intel compute stick) to improve the detection speed. The rPi 5 seems to be even faster, but I was curious if you have tried those compute sticks?
Good video. btw I had exactly the same problem with some eSun PLA+
Thanks for letting me know this - it’s definitely put me off ordering that brand again
Hello Kevin, thanks for the interesting video! Is the hand detector scale invariant? If it is, I am thinking of detecting other objects, not hands. Do I have to train them at first?
Yes you’ll need to train for specific objects - opencv has pre-trained models for faces and hands
i have an arducam 64 mp hawkeye autofocus camera working on bookworm lite 64 bit. Any chance you can make a tutorial for this with those? or do you know if it would work if on bookworm instead of on rasbian OS?
Great video! Love your fan girl!
can you use it to detect spicific face and connect to bt device/turn off bt?
does this work with raspberry pi HQ camera?
I am running some project with Image processing using raspbberry pi earlier. But I'm a little bit confuse about should I buy a brand new raspberry pi 5 for my project or its okay to just buy a rp 4 coz the budget limit that I have. Is anyone can help me ?
wre do is the code. I want test it but the code is missing
How does this work, Pi 5 with openCV doesn't work with the command cv2.VideoCapture(0), I know this for a fact. Is there a solution to this problem? Because there is nothing in the notes or instructions. Everyone who tries this code on a Pi5 will get an error and a blank video feed. Please help I'm confused.
The same problem happened to me and it's not about Pi 5, the problem is in the new OS (bookworm), OpenCV is working only in the old OS (bullseye) which is not available in Pi 5, i think in OS 32 Legacy only. Please tell me if you found a solution ?
I have the same issue/problem (14 Mar 2024) with the code supplied running on a new Pi5 and the OS Debian GNU/Linux 12 (bookworm)
anyone who gets past this please share some info.
OpenCV supports V4L/V4L2. That is the standard for video devices on Linux.
“Pi cameras” refuse to support this. That is Raspberry Pi’s problem, not OpenCV’s problem. Found it on the OpenCV site
Does this work on Raspberry Pi 5 with Bookworm? I'm very new to Raspberry Pi, trying to limp along but I don't think I'm installing CVZone correctly.
Oh yes - definitely should work on Pi 5 with Bookworm. Are you using a virtual environment - Bookworm pretty much prefers this way of using Python3 so that maybe why there is an issue?
@@kevinmcaleer28 Sorry for the dumb question, but I am trying to get this to work on a stock/new Raspberry pi 5 with Bookworm I am new to much of this.. Do you have or could you make step-by-step best install/setup guide for a Noob using a RP5 with Bookworm? Im not sure how to use the virtual environment. I would LOVE to be able to do some of the face tracking exactly as you show in your video.. Not 100% how to install any of this correctly. Thank you.
And how fast is it in FPS?
Hi tried so many times but without good results it came error no module name cvzone why?
What method did you use to install cvzone?
@@kevinmcaleer28 I made environment and made all the installation there but the cvzone is installed normally python I have 3.10 .just brand new pi 5 with the latest fw.
@@kevinmcaleer28 my problem is installing normally with pip install and came the same error
Good information but the thumbnail lead me to believe that you are showing how to train the pi to recognise certain people when a face is detected but alas.
@@ryanstacey2296 that's called facial recognition which is different from face detection
Where is the code
Surely it should detect you as a Java enthusiast?
Haha, however it only detects the Truth
@@kevinmcaleer28 maybe it's your subconscious truth! Twitter really does get silly sometimes.
Java for the win!