0
$\begingroup$

I'm trying to use AI and computer vision techniques to identify scene changes for a camera. Something like this:

enter image description here

enter image description here

What are some approaches to do this? Any ideas?

The scene is static. Somewhere I saw a user trains the app with multiple images of normal scenes, and the app monitors the amount of pixel change from the trained images. And it issues an alarm when the amount of pixel change exceeds a threshold for a specified time period. But it should be more complex than that. A human may walk around or shadows and sunlight might affect the scene, so it won't be a scene change and no anomaly.

Improve this question
$\endgroup$
4
  • 1
    $\begingroup$ If you have only such static scene and want to know if a static object where removed, you probably dont dont need deep learning. Just do background subtraction. $\endgroup$
    – Lelouch
    Commented Jul 12, 2023 at 18:58
  • $\begingroup$ It is static; but remember we have shades, shadows and humans passing. Does it work? Actually a simpler approach would be better! $\endgroup$
    – Mary
    Commented Jul 12, 2023 at 19:03
  • $\begingroup$ What they are suggesting is literally do a difference on the two images. The pixels will be different and you will know that something has changed. $\endgroup$
    – Ted Taylor of Life
    Commented Sep 15, 2023 at 11:10
  • $\begingroup$ crosspost: forum.opencv.org/t/door-blockage-detection-using-opencv/13927 $\endgroup$
    – Christoph Rackwitz
    Commented Sep 30, 2023 at 15:11

0

Reset to default

You must log in to answer this question.