Find Dominant Colors in an image

Contains two versions:

1. a cpp command line utility
2. an Objective-c Category that extends UIImage to find its dominant colors.

Overview

This demonstrates how to find the dominant colors in an image. Included are Swift and command-line(cpp) samples. The command-line version outputs a PNG file of the palette and a PNG image quantized to the resulting palette. The Swift version uses an Objective-C Category to extend UIImage and Objective-C Bridging Header to expose the extension to Swift. A Camera preview allows the user to take a photo from which dominant colors are found.

Based on algorithms found here

Requirements

• OpenCV 3.0 -- OpenCV is included as a submodule. The opencv2.framework needs to be built for ios.
• pkg-config -- used to find the necessary libs when building the command line utilty

Building the Swift app

1. This project has the OpenCV repo as a submodule.

   • If you want to use it be sure to do git submodule init. The opencv2.framework will be built by an XCode build script
   • If you wish to use your own (probably because the framework takes so long to build and opencv2.framework is already built) then open opencv.xcconfig and edit the path to point to your own clone of the opencv repo at https://github.com/opencv/opencv

2. Select your device in the device list and hit ⌘R

Running the command line:

• use the included makefile to compile the command line version

./getDominantColors <image> <number of colors>

• image is the image your wish to quantize
• the number of colors is the number of dominant colors you wish to find.