基于启发式模拟退火的多边形近似算法

将曲线多边形近似问题转换为最小化代价函数的问题,利用模拟退火算法对其求解最优解,并采用启发式方法将曲线的局部特征作为先验知识引入退火过程加速其收敛。通过实验比较表明,该方法在数据压缩率和近似误差等方面具有更好的性能,且压缩了运行时间。

Optimal Polygon Approximation Of Planar Curve Using Heuristic Simulated Annealing

The problem of polygon approximation was formulated as one of minimizing cost function. The optimal solution was found by using simulated annealing algorithm, and local features was introduced as prior knowledge into the algorithm in heuristic manner to accelerate its convergence. Compared with experimental results, the better effect and shorter running time were got by the algorithm proposed.