Getting Started with the OpenMV Cam: Machine Vision with MicroPython | Digi-Key Electronics
ฝัง
- เผยแพร่เมื่อ 10 ก.ย. 2024
- In this tutorial, Shawn shows you how to get started with the OpenMV H7 camera module, which uses MicroPython to perform a variety of machine vision tasks. Additionally, he provides a demonstration of person detection using the built-in TensorFlow Lite machine learning library.
The OpenMV IDE can be downloaded from this link: openmv.io/page...
Once installed, the IDE will start with a “Hello, World!” application. If you connect the IDE to your OpenMV camera, it will likely walk you through the process of updating the camera’s firmware. Once connected, you can run the default program to see what the camera sees.
In the IDE, select File - Examples - Machine Learning - tf_person_detection_search_whole_window to view the person detection example. In the video, we modify the example code to flash an LED whenever a person (or person-like figure) enters the field of view of the camera.
The demo program relies on the TensorFlow Lite for Microcontrollers module to perform inference using a pre-trained machine learning model. We load the model in code and then call the .classify() method, passing it the image taken from the camera. The image is classified as containing a person, no person, or unknown.
If the probability that a person is in the frame is over 0.7 (our set threshold) we turn on the LED.
Documentation for the OpenMV API can be found here: docs.openmv.io/
Product Links:
OpenMV H7 Camera - www.digikey.co...
OpenMV H7 Camera PLUS - www.digikey.co...
Related Videos:
What is Edge AI?
• Intro to Edge AI: Mach...
TinyML: Getting Started with TensorFlow Lite for Microcontrollers
• TinyML: Getting Starte...
Related Project Links:
Getting Started with the OpenMV Cam: Machine Vision with MicroPython:
www.digikey.co...
Related Articles:
What is Edge AI? - www.digikey.co...
TinyML: Getting Started with TensorFlow Lite for Microcontrollers - www.digikey.co...
is it me only or everyone is getting the perfect video of shawn on everytopic everytime😇
I really enjoy every time I watch Shawn videos...
Thanks!
I am just getting started with this device. I am glad Digi-Key has this type of video tutorial. Thanks
Shawn! Never a better way to start a new machine vision project!
Shawn, I too enjoy watching and learning everything I can from your channel. This episode has the makings of a second security system that I will create this time. 10 years ago I could hear and see the need for home security so I installed a eight camera system around my house. My neighbors though I was being paranoid, but as time moved on and attitudes changed they too could see the need. My neighbors have started installing those doorbell camera systems but they are extremely limited. With your camera ideas I think I will install a remote system on my mailbox that I can be powered with batteries and a solar panel. I may incorporate some PIR sensors to activate the cameras and limit battery usage. This is going to be a trial and error process in determining power requirements??? Thanks for a great episode because these commercial systems eat up monies every month just to give their customers what they should be provided. Facial recognition maybe difficult during these mask wearing times but hopefully this pandemic won’t last forever! Fingers crossed 🤞 take care!
Sounds like a great project! Let us know how it turns out 😀
Shawn Hymel - I just ordered the components and will need to determine power requirements then design & build batter and solar system to match with an average of 25% over watt production to allow for power production changes over the year.
I am getting an OS error of could not find the file. m using openMV IDE 2.9.7.
Same. U find a solution?
Excellent video. Very instructive!!!!!
Is it possible to "test" the camera before buying it by emulating videos or link the IDE to a phone for example ?
where can I get the model, I can't find this example in my IDE. I've tried the vww_96_int8 model but got a memory allocating error so I want to get your model to check whether I deployed incorrectly or not.
Thank you
Hi! Thanks a lot.
I want the using the esp32 camera via micro Python.But I could not find the useful data on the net.
Could you help me about that project?
In that project esp32 board is not connected the pc so esp32 board runs undependent.
When esp32 cam caught the red matter or blue matter I will control the LED.
Thanks a lot.
I have these cameras and I have used them a bit. There's a TON of arduino support for this, but I can understand the desire to use python instead. Have you had any luck finding python support for this?
Can I count people in the room all day with this camera? Is it possible to write such a program? Thank you for your knowledge, greetings from Poland
Excellent video indeed :)
This is a great video! Can we post this to our website?
Please do! That would be great!
@@ShawnHymel Done! It's on our homepage!
@@OpenMVLLC Wow, thank you! :D
I'm so distracted by the moire on his shirt ^^°
very nice
Thank you for such an illustrative video!
Can I import my own created neural network?