Computer vision is the extraction of information from either still or moving images. For instance, the information could be used in a facial recognition application to identify a specific individual in a crowd, and using the surrounding scene, the time and place the individual was present. Or it could be used to identify parts in a robotic assembler.

Some of the material in this section is adapted from Modern Algorithms for Image Processing by Vladimir Kovalevsky.

We are interested here in the following aspects of Computer Vision:

Shape recognition

• Recognition of circular objects
• From angle
• From shading

Noise reduction

• Spike noise
• White noise

Contrast enhancement

• Manual enhancement
• Automatic enhancement

Shading Correction

Edge Detection

• Basic theory
• Differential gradient
• Hough transform

Image compression

Image segmentation

3-D Vision and Motion

We begin by defining a C# class called CImage and two constructors:

  class CImage
  {
    public Byte[] Grid;
    public int width, height, nBits;
    // default contructor
    public CImage(){}
  }
  public CImage(int nx, int ny, int nbits)
  {
    width = nx;
    height = ny;
    nBits = nbits;
    Grid = new byte[width*height * (nbits/8)];  
  }