I'm using USB camera for object detection. I can't find code for them because the common blogs or articles are about object detection using pi camera only. Is object detection possible with a USB camera?

  • pyimagesearch.com/2017/10/16/…, github.com/EdjeElectronics/…, dlology.com/blog/… - first three results when googling "object detection with raspberry pi usb camera". All three seem like pretty good options for getting started with object detection. Sep 2 '19 at 20:05
  • I found another video explaining why any generic camera can do tensorflow object detection. It explains how you can use any camera to output video to Rpi as a series of jpeg images at 30 frames per second. Rpi checks out frame by frame using tensorflow to detect the object of interest. The tutorial uses the "docker" thing which means very easy "installation". Also the tutorial explains how to use Rpi python and free (well, for limited but enough newbie trials) cloud tensor flow API. I found it very good for lazy tensorflow newbies like me, not wanting to do any installation.
    – tlfong01
    Sep 3 '19 at 12:03
  • This is the youTube I recommended above. "Detect ANY Object with Raspberry Pi and TensorFlow" youtube.com/watch?v=zqIBce4LKx8. I forgot to mention that even your know middle level maths and physics, you need to spend perhaps two hours or more to google to learn tensorflow basics, thought google's tensorflow tutorials already try to be newbie friendly.
    – tlfong01
    Sep 3 '19 at 12:06

I'll turn my comment into an answer: Yes, it certainly is possible. I've read through a couple guides and my personal choice would be this Github. The readme provides a pretty good tutorial on how to get object detection working with TensorFlow on a Raspberry Pi. There are a few special instructions for users with a USB webcam instead of a Pi Camera, but they are made very clear in the guide. If reading guides aren't really your thing, there's also a YouTube guide over the same project.

This site is temporarily in read only mode and not accepting new answers.

Not the answer you're looking for? Browse other questions tagged .