Textures with Natural Complexity

(Computer Graphics Lecture by Kenji Igarashi, the Master of Information Science)

• Creating textures with natural complexity
  • Such as mountain surfaces or grasslands
• Pre-Stored Texture
  • A simple method of preparing and loading images
  • Becomes difficult with larger sizes due to storage and loading processes
• Procedural Texture
  • Using noise functions (which appear random and scattered)
    • Noise functions only return discrete values
    • By connecting the points on a graph, we can create noise waves
  • Generating images with noise waves of different frequencies and amplitudes
    • 
    • Higher frequencies (points on the noise function closer together) create finer details in the noise
    • Overlaying multiple frequencies creates a fractal-like effect (small noise within larger noise)
• Texture Synthesis
  • Preparing small textures in advance
  • However, generating larger textures from these small sample textures
    • Similar to image completion?
    • Like using Photoshop for completion