What it does OpenCV for Processing wraps the official OpenCV Java API (version 2.4.5) in a library that feels native to the Processing environment. It handles the boring parts—JNI bindings, native library loading, type conversions—so you can call findFaces() or findEdges() instead of wrestling with Mat objects and CascadeClassifier boilerplate.

The interesting bit The library doesn’t trap you in its wrapper. You can drop down to raw OpenCV Java calls whenever you need something exotic, a pattern borrowed from Kyle McDonald’s ofxCv for OpenFrameworks. It’s training wheels that come off without a wrench.

Key highlights

• Face detection, edge detection (Canny/Scharr/Sobel), Hough lines, contour finding, background subtraction, and histogram operations
• Region-of-interest support for processing image subsections
• Color channel splitting (RGB and HSV) with hue-based object detection
• Stereo depth calculation and camera calibration examples for advanced use
• Cross-platform: Mac OS X, 32/64-bit Windows, 32/64-bit Linux

Caveats

• Android support is “hopefully coming soon” (direct quote from 2013-era README)
• Live camera access broken on Retina Macs in Processing 2.0
• Some examples (CalibrationDemo, WarpPerspective) marked as “in progress”
• Based on OpenCV 2.4.5, which is now several major versions behind

Verdict Grab this if you’re teaching creative coding or prototyping vision ideas in Processing and want results before lunch. Skip it if you need modern OpenCV features, mobile support, or production-grade performance—you’ll want native bindings or a newer wrapper.