Scholarlink

<p>Floatplanes are a staple mode of transportation in remote places around the world including Alaska. Their rudders are a known vector for transmitting aquatic vegetation such as <i>Elodea</i> spp. between water bodies. There is limited research in understanding the details of transmission or on in-flight monitoring of entangled vegetation. Our study used strut-mounted GoPro cameras on two different model floatplanes to quantify vegetation transmissions for the duration of the flight. We developed and evaluated multiple computer vision models to automatically detect entangled vegetation and used these data to study transmission. The final model detected vegetation with 83.9% accuracy and an <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="10530_2025_3608_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="18" /> </InlineMediaObject> <EquationSource Format="TEX">\(F_1\)</EquationSource> <EquationSource Format="MATHML"><math> <msub> <mi>F</mi> <mn>1</mn> </msub> </math></EquationSource> </InlineEquation> score of 0.82. Additional analysis shows that of the 70 flights initially contaminated with vegetation, 60 (85.7%) still had entangled vegetation at landing. Our computer vision model is deployable on tablet computing devices commonly found in floatplanes, meaning a real-time detection and alert system could be developed to mitigate the further spread of invasive aquatic species.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Detecting aquatic plant transmission on floatplanes using computer vision

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月22日

查看全文
  • Patrick Pragman,
  • Frank D. W. Witmer,
  • Tobias Schwoerer

摘要

Floatplanes are a staple mode of transportation in remote places around the world including Alaska. Their rudders are a known vector for transmitting aquatic vegetation such as Elodea spp. between water bodies. There is limited research in understanding the details of transmission or on in-flight monitoring of entangled vegetation. Our study used strut-mounted GoPro cameras on two different model floatplanes to quantify vegetation transmissions for the duration of the flight. We developed and evaluated multiple computer vision models to automatically detect entangled vegetation and used these data to study transmission. The final model detected vegetation with 83.9% accuracy and an \(F_1\) F 1 score of 0.82. Additional analysis shows that of the 70 flights initially contaminated with vegetation, 60 (85.7%) still had entangled vegetation at landing. Our computer vision model is deployable on tablet computing devices commonly found in floatplanes, meaning a real-time detection and alert system could be developed to mitigate the further spread of invasive aquatic species.