Pynopticon is a toolbox that allows you to create and train your own object recognition classifiers. It makes rapid prototyping of object recognition work flows a snap. Simply create a dataset of your favorite image categories, choose some feature extraction methods (e.g. SIFT), post-processing (clustering, histograms, normalization) and a classifier to train (e.g. SVM, Bayes, Descision Trees) and leave the rest to Pynopticon. Pynopticon integrates with the Orange toolbox for machine learning which provides it with a userfriendly GUI. Special care has been taken to let the user decide if he wants to have a lower memory footprint or be computationally more efficient. All this makes Pynopticon a perfect tool for lectures and their students, beginners in object recognition but also computer vision researches who can easily integrate their own algorithms into Pynopticon.
- Changes to previous version:
Initial Announcement on mloss.org.
No one has posted any comments yet. Perhaps you'd like to be the first?
Leave a comment
You must be logged in to post comments.