关键词: 特征点, B样条曲线, 粒子群算法, 曲线逼近, 逆向工程

Abstract: To reduce the control points of B-spline curve used in contour reconstruction of reverse engineering,a B-spline curve approximation algorithm based on feature point extraction and Particle Swarm Optimization (PSO) was proposed.Equal chord length method was used to calculate the curvature of the discrete points,and an initial curve was constructed with the use of feature points extracted by curvature analysis.One new interpolation point was inserted at the place where the maximum fitting error occurs to construct a new approximate curve.This process was repeated until the fitting error wasnt greatly reduced when a new interpolation point was inserted.An improved PSO algorithm was applied to optimize the position of control points to get the final B-spline curve.The experimental result showed that the proposed algorithm could reduce more control points with high efficiency and accuracy by comparing with other methods.

Key words: feature points, B-spline curve, particle swarm optimization, curve approximation, reverse engineering