OpenCV for Processing

A Processing library for the OpenCV computer vision library.

OpenCV for Processing is based on OpenCV's official Java bindings. It attempts to provide convenient wrappers for common OpenCV functions that are friendly to beginners and feel familiar to the Processing environment.

See the included examples below for an overview of what's possible and links to the relevant example code. More comprehensive documentation coming soon.

OpenCV for Processing is based on the officially supported OpenCV Java API, currently at version 2.4.5. In addition to using the wrapped functionality, you can import OpenCV modules and use any of its documented functions: OpenCV javadocs. See the advanced examples (HistogramSkinDetection, DepthFromStereo, and Marker Detection) below for details. (This style of API was inspired by Kyle McDonald's ofxCv addon for OpenFrameworks.)

Contributions welcome.

Installing

OpenCV for Processing currently supports Mac OSX, 32-bit and 64-bit Windows, 32- and 64-bit Linux. Android support is hopefully coming soon (pull requests welcome).

NB: When running on the Mac, make sure you have Processing set to 64-bit mode in the Preferences

See here for the latest release.

Examples

LiveCamTest

Access a live camera and do image processing on the result, specifically face detection.

Code: LiveCamTest.pde

Note: There's a bug that prevents live camera access in current versions of Processing 2.0 on machines with a Retina display.

FaceDetection

Detect faces in images.

Code: FaceDetection.pde

BrightnessContrast

Adjust the brightness and contrast of color and gray images.

Code: BrightnessContrast.pde

FilterImages

Basic filtering operations on images: threshold, blur, and adaptive thresholds.

Code: FilterImages.pde

FindContours

Find contours in images and calculate polygon approximations of the contours (i.e., the closest straight line that fits the contour).

Code: FindContours.pde

FindEdges

Three different edge-detection techniques: Canny, Scharr, and Sobel.

Code: FindEdges.pde

FindLines

Find straight lines in the image using Hough line detection.

Code: HoughLineDetection.pde

BrightestPoint

Find the brightest point in an image.

Code: BrightestPoint.pde

RegionOfInterest

Assign a sub-section (or Region of Interest) of the image to be processed. Video of this example in action here: Region of Interest demo on Vimeo.

Code: RegionOfInterest.pde

ImageDiff

Find the difference between two images in order to subtract the background or detect a new object in a scene.

Code: ImageDiff.pde

DilationAndErosion

Thin (erode) and expand (dilate) an image in order to close holes. These are known as "morphological" operations.

Code: DilationAndErosion.pde

BackgroundSubtraction

Detect moving objects in a scene. Use background subtraction to distinguish background from foreground and contour tracking to track the foreground objects.

Code: BackgroundSubtraction.pde

WorkingWithColorImages

Demonstration of what you can do color images in OpenCV (threshold, blur, etc) and what you can't (lots of other operations).

Code: WorkingWithColorImages.pde

ColorChannels

Separate a color image into red, green, blue or hue, saturation, and value channels in order to work with the channels individually.

Code: ColorChannels

FindHistogram

Demonstrates use of the findHistogram() function and the Histogram class to get and draw histograms for grayscale and individual color channels.

Code: FindHistogram.pde

HueRangeSelection

Detect objects based on their color. Demonstrates the use of HSV color space as well as range-based image filtering.

Code: HueRangeSelection.pde

CalibrationDemo (in progress)

An example of the process involved in calibrating a camera. Currently only detects the corners in a chessboard pattern.

Code: CalibrationDemo.pde

HistogramSkinDetection

A more advanced example. Detecting skin in an image based on colors in a region of color space. Warning: uses un-wrapped OpenCV objects and functions.

Code: HistogramSkinDetection.pde

DepthFromStereo

An advanced example. Calculates depth information from a pair of stereo images. Warning: uses un-wrapped OpenCV objects and functions.

Code: DepthFromStereo.pde

WarpPerspective (in progress)

Un-distort an object that's in perspective. Coming to the real API soon.

Code: WarpPerspective.pde

MarkerDetection

An in-depth advanced example. Detect a CV marker in an image, warp perspective, and detect the number stored in the marker. Many steps in the code. Uses many un-wrapped OpenCV objects and functions.

Code: MarkerDetection.pde