Feather Animation Based on an Intelligence Algorithm

Abstract

A new approach for the simulation of feather based on a statistical method is presented. The feather boundary can be firstly detected according to its configuration, and then its contour shape can also be simulated on the basis of a curve-fitting technique. The components within its boundary are divided into two parts, i.e. the heavier one and the softer one. The former region may be filled with the black and white patterns depending upon the necessary information detected before. And the latter region may be represented by the two parts, i.e. the softer rachis and down. Distribution probability of the rachis curve shape is derived from the detected information so that the simulation can be realized. The softer down may be represented by the statistical method after efficiently analysis its 2D gray image information.