论文部分内容阅读
本文证明了平均绝对差地形匹配算法失配度量的渐近分布是正态分布,提出了计算代价极小化的数值算法。研究了计算代价多级优化的方法和对于进一步减小计算量的必要性。讨论了基准数字地图局部零均值化预处理的方法和应用价值。大量的计算机仿真结果证实了理论分析和所研究方法的正确性,使实际匹配时间降低到常规平均绝对差算法的1/6。
This paper proves that the asymptotic distribution of the mismatch measure of the mean absolute difference terrain matching algorithm is a normal distribution and proposes a numerical algorithm that minimizes the computational cost. The multi-level optimization method of computational cost and the necessity of further reducing the computational cost are studied. This paper discusses the method and application value of local zero-mean-value preprocessing of reference digital map. A large number of computer simulation results confirm the correctness of the theoretical analysis and the research methods, and reduce the actual matching time to one sixth of the conventional average absolute difference algorithm.