Practical Computer Vision with SimpleCV

Chapter 6. Image Arithmetic

As discussed in Chapters 3 and 4, we can think of images as a grid of pixels with each pixel having a color defined as an RGB triplet. Each component in that RGB triplet has a value in the range of 0 to 255. Given this structure and format, it is actually quite easy to perform arithmetic on images, such as addition, subtraction, multiplication, and division. This concept should not be entirely new. Chapter 5 already introduced this idea when using color to segment images. The goal of this section is to examine this concept in further detail, and understand how mathematical manipulation of images is used in a vision system. Topics include:

• The basic mathematical operations from elementary school, as applied to images

• Using histograms to calibrate the camera

• Finding the dominant colors of an image and using that information to segment the image

Basic Arithmetic

Just as with elementary school arithmetic, the easiest place to start is with addition. Two images can be added together. Underneath the hood, the SimpleCV framework goes through the two images pixel by pixel, and adds the RGB values of the pixels at corresponding locations together. Starting at pixel (0,0) on both images (the top left corner of each image), it will add the values for each component of the RGB triplet together to get a new RGB triplet. For instance, if the RGB value of the pixel at (0,0) in the first image is (1,2,3), and the RGB value at the same (0,0) point in the second image is