Scholarlink

<p>In Mexico, dengue is one of the most critical public health problems. Entomological surveillance is essential for the selection of vector control methods. Ovitraps are used for entomovirological surveillance, identifying areas with vertical transmission in <i>Aedes</i> eggs. One of the principal uses of ovitraps consists of counting eggs to construct operational indexes that allow estimating mosquito populations and risk models and contribute to identifying priority areas for entomological surveillance. This paper aims to develop an algorithm that automatically counts mosquito eggs from images. We create the EggCountATT algorithm based on adaptive thresholding and a thinning operator. EggCountATT outperforms the methods ICount, MECVision, and EggCountAI, which are evaluated in two datasets. EggCountATT reaches a mean accuracy of 92.09% in our dataset, which contains images with several egg clusters and overlapping eggs.</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

Aedes aegypti mosquito egg-counting using adaptive thresholding and thinning

  • 期刊论文

  • 发表日期：

    2026年2月17日

查看全文
  • Francisco J. Hernandez-Lopez,
  • Kenia Mayela Valdez-Delgado,
  • Victor Muñiz-Sánchez,
  • Cuauhtémoc Villarreal-Treviño,
  • Graciela González-Farías,
  • Rogelio Danis-Lozano

摘要

In Mexico, dengue is one of the most critical public health problems. Entomological surveillance is essential for the selection of vector control methods. Ovitraps are used for entomovirological surveillance, identifying areas with vertical transmission in Aedes eggs. One of the principal uses of ovitraps consists of counting eggs to construct operational indexes that allow estimating mosquito populations and risk models and contribute to identifying priority areas for entomological surveillance. This paper aims to develop an algorithm that automatically counts mosquito eggs from images. We create the EggCountATT algorithm based on adaptive thresholding and a thinning operator. EggCountATT outperforms the methods ICount, MECVision, and EggCountAI, which are evaluated in two datasets. EggCountATT reaches a mean accuracy of 92.09% in our dataset, which contains images with several egg clusters and overlapping eggs.