An addon for capturing sets of keypoints and descriptors and then finding one or more objects in an image.
Latest commit 5bd6063 Jan 17, 2014 @jbeuckm Update
Failed to load latest commit information.
docs instructions Jan 17, 2014
example-ofxUI instructions Jan 17, 2014
src dummy object name for new ones Jan 14, 2014
.gitignore initial commit Dec 22, 2013 Update Jan 17, 2014 config Jan 14, 2014 initial commit Dec 22, 2013
ofxaddons_thumbnail.png thumbnail resized Jan 16, 2014



Analyze video grabber frames (or a loaded image) for SURF keypoints. Draw a lasso around the keypoints you would like to use in a detector. Then save object descriptor files. Later, load descriptor files and detect objects in new images.

  • Choose an object that generates lots of features. Textures, patterns and text are good candidates for feature detection.

Find Features

  • Circles show the rough size of each detected feature. Gray features are the weakest and white are the strongest. The more good features, the better a detector.

Find Features

  • Draw a polygon around the features you want a detector to use.

Find Features

  • Save the object descriptors to a YML file. After saving, and whenever you load the descriptor file, the detector will try to find your object in the latest frame.

Find Features





Known issues

Version history

This addon is under development, but can be used as-is.