Streaming camera frames for template matching

Template matching is a machine-learning technique to find areas of an image that match a given template image. We will apply template matching to every frame of a real-time video stream to locate an image.

Getting ready

Install the OpenCV and c2hs toolkits:

$ sudo apt-get install c2hs libopencv-dev

Install the CV library from cabal. Be sure to include the –fopencv24 or –fopencv23 parameter depending on which version of OpenCV is installed:

$ cabal install CV -fopencv24

Also, create a small template image. In this recipe, we use an image of Lena, which is usually used in many image-processing experiments. We name this image file lena.png:

How to do it…

In a new file, Main.hs, start with these steps:

1. Import the relevant libraries:

   {-#LANGUAGE ScopedTypeVariables#-}
   module Main where
   import CV.Image (loadImage, rgbToGray, getSize)
   import CV.Video (captureFromCam, streamFromVideo)
   import Utils.Stream (runStream_, takeWhileS, sideEffect...