Customizing your embedded system!

Now that you have created a whole embedded Cartoonifier system, and you know the basics of how it works and which parts do what, you should customize it! Make the video full screen, change the GUI, change the application behavior and workflow, change the Cartoonifier filter constants or the skin detector algorithm, replace the Cartoonifier code with your own project ideas, or stream the video to the cloud and process it there!

You can improve the skin detection algorithm in many ways, such as using a more complex skin detection algorithm (for example, using trained Gaussian models from many recent CVPR or ICCV conference papers at http://www.cvpapers.com), or add face detection (see the Face detection section of Chapter 5, Face Detection and Recognition with the DNN Module) to the skin detector, so it detects where the user's face is, rather than asking the user to put their face in the center of the screen. Be aware that face detection may take many seconds on some devices or high-resolution cameras, so they may be limited in their current real-time uses. But embedded system platforms are getting faster every year, so this may be less of a problem over time.

The most significant way to speed up embedded computer vision applications is to reduce the camera resolution absolutely as much as possible (for example, 0.5 megapixels instead of 5 megapixels), allocate and free images as rarely as possible, and perform image format conversions as rarely as possible. In some cases, there might be some optimized image processing or math libraries, or an optimized version of OpenCV from the CPU vendor of your device (for example, Broadcom, NVIDIA Tegra, Texas Instruments OMAP, or Samsung Exynos), or for your CPU family (for example, ARM Cortex-A9).