作者: Lu Jingxuan , Li Lin , Leng Wanchun , Li Xiaotao , Yang Yongmin
DOI:
关键词: Confusion matrix 、 Vector space 、 Computer science 、 Block (data storage) 、 Sample (graphics) 、 Set (abstract data type) 、 Grid 、 Scale (descriptive set theory) 、 Product (mathematics) 、 Data mining
摘要: The invention discloses a drawing precision evaluation method for large-scale surface water body product. comprises the following steps of firstly, collecting publicly published and globally covered products; selecting seasonal product capable representing dynamic information body; dividing into different block objects by adopting based on space grid constraint; re-classifying each object three vector subsets permanent body, non-water randomly generating sample points subset region to form set, verifying set combining Google earth high-resolution image expert knowledge experience; utilizing verified verification constructing confusion matrix, calculating global index Kappa coefficient product, thereby realizing