Image Subtraction using frame history [closed]
Hello Everyone,
I am trying to use image subtraction for extracting contours from the video. I just need to subtract the average of the first 20 frames for this particular problem. However, I figured out that current algos substract on a rolling image basis.
My current approach is: 1. Adding first 20 frames and then dividing by 20 to get average of the background 2. Substracting the averaged image from frames >20.
Is there any other way ?
Any help would be greatily appreciated ?
trying to reinvent background subtraction ?
What does that means ? Not trying to invent anything though
oh apologies, i'm more trying to pull your leg here.
but we do have some backgound subtraction implementations here already. have a look ! maybe you want to refine your question, then.
cool! Thanks Brother! I had a look at the link in the previous comment. I think what I want to ask is that for the problem that I face is that:
I want my background image to be constant instead of using the image on rolling basis that is using the previous image for current frame.
I am very kind of naive to image analysis so excuse my lingo.
Thanks Andy
ok, i see. but do you really have a "static" background ? if you have natural lighting, the sun will move, clouds will come & go. someone will move a chair, and it stays like that for the rest of the day.
I do have a static background as the video is slow motion video with 30 fps from the controlled engineering test performed under static lightening conditions.
see https://answers.opencv.org/question/5151