Skip to content

An implementation of a video stablizer based on feature point detection

Notifications You must be signed in to change notification settings

Isawan/Stabilizer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

78 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Stabilizer

A video stabilizer implementation based on feature point matching by Isawan Millican and Christopher Cole. The whole project from research, implementation to writing the report was completed in just under two weeks.

Here's an example of the program working.

Stabilized video

The software has been extended to allow image stiching from videos.

Stiched images

Make sure you have the dependencies installed:

  • numpy
  • matplotlib
  • skvideo
  • opencv
  • shapely

So let's start off by generating the mache image. Note -v is optional and assumes 30 fps.

python3 -m stabilizer.combine -i <input video> -f <output image> -v <start time in secs>:<end time in secs>

Generate the in-frame image

python3 -m stabilizer.stitch -i <input video> -f <output video>

Generate the normal stabilized video (Note: -fm is optional)

python3 -m stabilizer.stable -i <input video> -f <output video> \
            -fm <matrix print> 

You'll also see compare.py. That file is used for debugging so you can probably ignore it but for completeness

python3 -m stabilizer.compare

About

An implementation of a video stablizer based on feature point detection

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published