Summary

In this chapter, we explored a number of interesting image processing effects. We used dodging and burning to create a black-and-white pencil sketch effect, explored lookup tables to arrive at an efficient implementation of curve filters, and got creative to produce a cartoon effect.

One of the techniques used was two-dimesional convolution, which takes a filter and an image and creates a new image. In this chapter, we provided the filters to get the results we wanted, but we don't always have the filters that are necessary to produce the results we want. Recently, deep learning has emerged, which tries to learn the values for different filters to help it get the results it wants.

In the next chapter, we will shift gears a bit and explore the use of depth sensors, such as Microsoft Kinect 3D, to recognize hand gestures in real time.