【摘要】 为了解决传统无控制DEM最小二乘匹配方法拉入范围小的问题,首先建立了一种基于匹配度之和最大的DEM匹配模型,然后结合DEM匹配的实际需要,对遗传算法的编码方案、初始种群生成、适应度函数、基本遗传操作和迭代终止条件等5个基本方面进行了设计;在此基础上,设计了基于遗传算法和最小二乘匹配相结合的无控制DEM匹配方法流程,最后采用模拟数据设计了三组试验方案,测试新模型及算法的性能,并分别与传统模型的遗传匹配方法和最小二乘匹配方法进行了比较。结果表明:新算法比最小二乘匹配方法具有更大的拉入范围,比遗传匹配方法具有更快的匹配速度和更高的匹配精度。更多还原

【Abstract】 The traditional DEM matching can be accomplished by the least squares matching without control points,but its application is limited in the poor pull-in-range.In order to solve this problem,this paper established a DEM matching model with maximizing the sum of matching-degree firstly,and then,designed the five basic aspects of genetic algorithm,including the coding scheme,the initial population generation method,the fitness function,the basic genetic operators and the iteration stopping criteria,combined with the practical demands of DEM matching.On this basis,a procedure of DEM matching algorithm without control points was designed,in which genetic algorithm and least squares were combined together.At last,the capabilities of proposed model and method were tested by designing three sets of experimentations with simulated data,and compared with the matching methods of genetic algorithm using the traditional model and the least squares matching method.The experimental results showed that the method could get larger pull-in-range than the least squares matching method,and match faster and more accuracy than the genetic matching method with the traditional model.更多还原