This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
If you have captured video frames from when the equipment is not in the scene, then a simple method is to subtract the captured background from the current frame and sum up any large pixel differences to detect when the scene has changed. You can require that the large changes not be at the border of the scene as well, so that you will only capture a frame when the equipment is fully in the scene. However, this technique is NOT very robust, as any lighting, noise, other objects, or video gain changes can ruin its effectiveness. OpenCV has more sophisticated background subtraction as seen [here].(https://docs.opencv.org/3.3.0/db/d5c/tutorial_py_bg_subtraction.html).
I recommend that you read up on the reference papers for MOG background subtraction and try it.
| | 2 | No.2 Revision |
If you have captured video frames from when the equipment is not in the scene, then a simple method is to subtract the captured background from the current frame and sum up any large pixel differences to detect when the scene has changed. You can require that the large changes not be at the border of the scene as well, so that you will only capture a frame when the equipment is fully in the scene. However, this technique is NOT very robust, as any lighting, noise, other objects, or video gain changes can ruin its effectiveness. OpenCV has more sophisticated background subtraction as seen [here].(https://docs.opencv.org/3.3.0/db/d5c/tutorial_py_bg_subtraction.html).[here] (https://docs.opencv.org/3.3.0/db/d5c/tutorial_py_bg_subtraction.html).
I recommend that you read up on the reference papers for MOG background subtraction and try it.
