Fitting of Continuous Line Segments with Arc Paths

In this paper, we propose a half-constraint least squares algorithm to fit the continuous line segments with arc paths. In particular, we develop a method to calculate the nodes, with which the fitting algorithm is more robust against the noise and the arc path contains more arcs. Moreover, the tolerance of the first-order derivative is considered for the nodes, so that the arc path becomes smoother. Numerical simulations and experiments are conducted to verify the validity of the proposed method and to show the improved performance in the sense of high precision.