科技创新与应用2025，Vol.15Issue(7)：85-88,4.DOI:10.19981/j.CN23-1581/G3.2025.07.021
基于特征组合的无人机影像城市地表要素提取
摘要
Abstract
Urban surface elements are the basic components of urban management and planning,and accurate acquisition of their information is of great significance for urban planning and social and economic analysis.Based on some drone image data in Jinzhai County,this paper uses an object-oriented classification method to study the extraction of urban surface elements,and explores the impact of the number of features and different feature combinations on classification results.The results show that,the selection of features is fully considered,and the more the number of features is selected,the higher the accuracy of the result;in urban feature extraction,the Bayesian model has higher accuracy than the KNN model,with the overall accuracy reaching 84%,and the Kappa coefficient is 0.78.关键词
无人机/城市地表要素/面向对象/特征优选/精度Key words
drone/urban surface elements/object-oriented/feature optimization/accuracy分类
天文与地球科学引用本文复制引用
宋素素,范丽丽,朱静..基于特征组合的无人机影像城市地表要素提取[J].科技创新与应用,2025,15(7):85-88,4.基金项目
滁州职业技术学院校级科研项目(ZKZ-2024-5、ZKY-2024-7) （ZKZ-2024-5、ZKY-2024-7）
