Suppr超能文献

无人机成像与实地巡查对比,以识别和分类作为埃及伊蚊潜在滋生地的垃圾场。

Comparison of unmanned aerial vehicle imaging to ground truth walkthroughs for identifying and classifying trash sites serving as potential Aedes aegypti breeding grounds.

作者信息

Tarpenning Morgan S, Bramante Juliet T, Coombe Kavita D, Woo Katherine E, Chamberlin Andrew J, Mutuku Paul S, De Leo Giulio A, LaBeaud Angelle Desiree, Ndenga Bryson A, Mutuku Francis M, Rosser Joelle I

机构信息

Stanford University, Stanford, CA, USA.

University of Washington, School of Medicine, Seattle, WA, USA.

出版信息

Parasit Vectors. 2025 Mar 6;18(1):93. doi: 10.1186/s13071-025-06706-1.

Abstract

BACKGROUND

Trash piles and abandoned tires that are exposed to the elements collect water and create productive breeding grounds for Aedes aegypti mosquitoes, the primary vector for multiple arboviruses. Unmanned aerial vehicle (UAV) imaging provides a novel approach to efficiently and accurately mapping trash, which could facilitate improved prediction of Ae. aegypti habitat and consequent arbovirus transmission. This study evaluates the efficacy of trash identification by UAV imaging analysis compared with the standard practice of walking through a community to count and classify trash piles.

METHODS

We conducted UAV flights and four types of walkthrough trash surveys in the city of Kisumu and town of Ukunda in western and coastal Kenya, respectively. Trash was classified on the basis of a scheme previously developed to identify high and low risk Aedes aegypti breeding sites. We then compared trash detection between the UAV images and walkthrough surveys.

RESULTS

Across all walkthrough methods, UAV image analysis captured 1.8-fold to 4.4-fold more trash than the walkthrough method alone. Ground truth validation of UAV-identified trash showed that 94% of the labeled trash sites were correctly identified with regards to both location and trash classification. In addition, 98% of the visible trash mimics documented during walkthroughs were correctly avoided during UAV image analysis. We identified advantages and limitations to using UAV imaging to identify trash piles. While UAV imaging did miss trash underneath vegetation or buildings and did not show the exact composition of trash piles, this method was efficient, enabled detailed quantitative trash data, and granted access to areas that were not easily accessible by walking.

CONCLUSIONS

UAVs provide a promising method of trash mapping and classification, which can improve research evaluating trash as a risk factor for infectious diseases or aiming to decrease community trash exposure.

摘要

背景

暴露在自然环境中的垃圾堆和废弃轮胎会积水,为埃及伊蚊创造了理想的滋生地,而埃及伊蚊是多种虫媒病毒的主要传播媒介。无人机成像提供了一种新颖的方法来高效、准确地绘制垃圾分布图,这有助于改进对埃及伊蚊栖息地及后续虫媒病毒传播的预测。本研究评估了通过无人机成像分析识别垃圾的效果,并与在社区中步行计数和分类垃圾堆的标准做法进行比较。

方法

我们分别在肯尼亚西部的基苏木市和沿海的乌昆达镇进行了无人机飞行和四种类型的实地垃圾调查。垃圾根据先前制定的用于识别埃及伊蚊高风险和低风险滋生地的方案进行分类。然后,我们比较了无人机图像和实地调查之间的垃圾检测情况。

结果

在所有实地调查方法中,无人机图像分析捕捉到的垃圾比单独的实地调查方法多1.8倍至4.4倍。对无人机识别的垃圾进行地面实况验证表明,94%的标记垃圾站点在位置和垃圾分类方面都被正确识别。此外,在无人机图像分析过程中,98%在实地调查期间记录的可见垃圾模拟物被正确避开。我们确定了使用无人机成像识别垃圾堆的优点和局限性。虽然无人机成像确实遗漏了植被或建筑物下方的垃圾,并且没有显示垃圾堆的确切组成,但这种方法效率高,能够提供详细的定量垃圾数据,并且可以进入步行难以到达的区域。

结论

无人机提供了一种很有前景的垃圾测绘和分类方法,可改善将垃圾作为传染病风险因素进行评估或旨在减少社区垃圾暴露的研究。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/2587/11883972/3cd5bd67cf7f/13071_2025_6706_Fig1_HTML.jpg

文献检索

告别复杂PubMed语法,用中文像聊天一样搜索,搜遍4000万医学文献。AI智能推荐,让科研检索更轻松。

立即免费搜索

文件翻译

保留排版,准确专业,支持PDF/Word/PPT等文件格式,支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述,25分钟生成高质量综述,智能提取关键信息,辅助科研写作。

立即免费体验