Simple tests on PGM based image processing

• write a program capable of read, write and generate simple pgm images
• write fuctions capable of manipulating these images and obtain data
  • noise reduction
  • geometric description of circles

• generate a noisy image of a circle
  • draw a circle by giving its parameters
  • apply noise filter
• generate and remove grayscale noise
  • normalization
  • binarization
• obtain multiple circle geometry
  • absolute position of center
  • relative positioning

Small image (50x50) with extent noise in proportion to the circle size

Bigger images usually require more than a simple rand() per pixel as noise algorithm. A more natural noise would be best described in parameters of different orders of magnitude as seem from further algorithms. Grouping occurence, color variation, deformation gradients.. An algorithm like Perlin's would help with this problem.