X-Git-Url: https://git.ao2.it/experiments/opencv_trail_effect.git/blobdiff_plain/d095022ca7b1d9eb46521073b71f2327fb9d5637..648650d64c6777240f5e9e8ba5b005e7b6f415c3:/README?ds=sidebyside diff --git a/README b/README index cd242b0..20ea8e5 100644 --- a/README +++ b/README @@ -9,17 +9,25 @@ Some examples of it are: The Jacksons - Blame It On The Boogie - https://www.youtube.com/watch?v=mkBS4zUjJZo OK Go - WTF? - https://www.youtube.com/watch?v=12zJw9varYE Elisa - L'Anima Vola - https://www.youtube.com/watch?v=MqhxIQD16EA + Kool & The Gang - Get Down On It - https://www.youtube.com/watch?v=qchPLaiKocI + Kool & The Gang - Steppin' Out - https://www.youtube.com/watch?v=KRuiVyq2Akc + Earth, Wind & Fire - September - https://www.youtube.com/watch?v=Gs069dndIYk opencv_trail_effect draws a trail after performing foreground segmentation -(using thresholding or background subtraction) using OpenCV. +using OpenCV. -The "sole" assumptions for a decent effect are that: +Segmentation in opencv_trail_effect can done using either thresholding or +background subtraction. - 1. the camera does not move; - 2. the background lighting in the scene is quite stable (when doing - background subtraction it is recommended to disable auto-gain in the - camera, and avoiding any artifact introduced by the power line - frequency). +A few conditions are required to create a decent effect: + + 1. the camera should not move; + 2. the lighting in the scene should be quite stable (when doing background + subtraction it is recommended to disable auto-gain in the camera, and + avoid any artifact introduced by the power line frequency); + 3. the subject should enter the scene only after the background learning + phase has finished, i.e. after opencv_trail_effect has shown the preview + window. Examples of use @@ -49,6 +57,8 @@ frames in the trail, possibly with the most recent frame drawn on top: $ ./opencv_trail_effect -l 10 -s background -d average -B +Here is a video showing the examples from above: +https://www.youtube.com/watch?v=X3oRozxt2O4 Side note -------------------