智能地物景像匹配定位技术研究

Intelligent scene\|matching positioning technology

文中研究了智能地物景像匹配定位系统的基本原理,从模式识别的观点出发提出了地物景像匹配算法的数学模型,并讨论了影响系统性能的有关因素。

The mechanism of intelligent scene\|matching positioning technology has systematically been studied. A mathematical model describing the scene\|matching algorithm is proposed from the view of pattern recognition. The factors affecting the positioning performance are theoretically analyzed using the model. A large number of computer simulations have been carried out with a normalized production correlation algorithm (NPROD). The results show that it can provide the intelligent scene\|matching positioning system with an accuracy (Circular Error Probable) of less than a cell in the referenced map and an acquisition probability of more than 97%. Therefore, NPROD can be used in the system as a basic scene correlation algorithm.