Mathematica Cookbook

Chapter 8. Image Processing

I have a picture Pinned to my wall An image of you and of me and we’re laughing We’re loving it all

...

You say I’m a dreamer We’re two of a kind Both of us searching for some perfect world We know we’ll never find

—Thompson Twins, “Hold Me Now”

8.0 Introduction

Image processing is a field with many challenges. The first challenge is the magnitude of the data. Consider that a simple 256 × 256 pixel grayscale image will contain 65,536 bytes of data for the pixel values alone. Larger color images can contain many times this amount. The second challenge is the raster form of the image data, which is optimized for display, not for detecting distinct visual elements. A third challenge is the noise and other artifacts of the image-capture process. A final challenge is the lack of contextual information; most images do not encode where they were taken, the lighting conditions, the device used, and so on (although this is beginning to change). In my opinion, these challenges make working on image processing very rewarding, especially when one considers that significant portions of our brains are dedicated to visual perception. Finding algorithms that achieve the kinds of visualprocessing tasks that the brain performs is one way to begin to peel away the veil obscuring the workings of our most mysterious organ.

The field of image processing is very broad; this chapter only samples a small fraction of the relevant problems. The choice of topics is largely a function ...