Stochastic reasoning for structural pattern recognition: An example from image-based UAV navigation
作者：
Highlights：
• By simulating unmanned air vehicle navigation a perception-action loop is closed.
• A virtual globe system is used as approximate open world.
• Recognition of landmarks (bridges) is performed by a knowledge-based system.
• EM estimation for assessing the recognition likelihood.
• Inlier/outlier mixture is insufficient—“intermediates” are required.
摘要
Highlights•By simulating unmanned air vehicle navigation a perception-action loop is closed.•A virtual globe system is used as approximate open world.•Recognition of landmarks (bridges) is performed by a knowledge-based system.•EM estimation for assessing the recognition likelihood.•Inlier/outlier mixture is insufficient—“intermediates” are required.
论文关键词：Structural pattern recognition,Decision rationales,Expectation–Maximization,Navigation,UAV
论文评审过程：Received 8 February 2013, Revised 8 November 2013, Accepted 18 February 2014, Available online 27 February 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.02.009
