Origami Project Goals
- Lowest barrier of entry to Image Processing, Computer Vision and Neural Networks on the JVM
- by using OpenCV (and when possible Clojure)
- Works out of the box on your device (OSX, Windows, Linux, Raspberry, Jetson ..) and in your language (Clojure, Java, Kotlin, Scala, Binary via Google's Bazel)
- Full support for OpenCV 4.9.0
- On-going support for Yolo v6
Origami Docs
- codox generated API is here
- Ready to use jupyter, clojure notebooks, java notebooks
- Deep Neural Networks with origami-dnn
- Run on AWS Lambdas
- Real-time Streaming Server App
- With origami-sources
- use various sources (folders, zip files, dropbox, webcam, flickr etc..) to lazily load OpenCV Mat objects
- use various handlers (You Tube, HTTPS etc...) to load VideoCapture objects


