Scholarlink

<p>Millions of people yearly suffer from respiratory issues due to airborne pollen, highlighting the urgent need for advanced pollen detection systems. This study presents a novel automated approach for pollen recognition using microscopy images, addressing the inefficiencies of traditional manual identification methods. The proposed system introduces three innovative pre-processing techniques: SigmaVision Deblurring, Micro-Denoise for Varied Noises (MDVN), and Quantum Contour Pollen Morph Segmentation (QCPMS). The methods developed were based on 2523 images representing 73 distinct species of pollen, and they improved image quality and segmentation accuracy by 25% and 19%, respectively. Utilizing DenseNet, the proposed deep learning model achieved an accuracy rate of 95.65%, outperforming traditional models such as ResNet. Beyond precise pollen classification, this technique has broader environmental applications, enabling accurate monitoring of pollen dispersal and contributing to allergy forecasting, biodiversity analysis, climate modeling, and public health initiatives related to pollen exposure. Ultimately, this technology lays the foundation for an integrated air quality monitoring and ecological research system, combining environmental models with advanced monitoring technologies.</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

PollenMorph AI: quantum contours based segmentation and deep learning for pollen recognition using microscopic images

  • 期刊论文

  • 发表日期：

    2025年4月2日

查看全文
  • Mahalakshmi Priya Rajendran,
  • Sumathi Mahadevan

摘要

Millions of people yearly suffer from respiratory issues due to airborne pollen, highlighting the urgent need for advanced pollen detection systems. This study presents a novel automated approach for pollen recognition using microscopy images, addressing the inefficiencies of traditional manual identification methods. The proposed system introduces three innovative pre-processing techniques: SigmaVision Deblurring, Micro-Denoise for Varied Noises (MDVN), and Quantum Contour Pollen Morph Segmentation (QCPMS). The methods developed were based on 2523 images representing 73 distinct species of pollen, and they improved image quality and segmentation accuracy by 25% and 19%, respectively. Utilizing DenseNet, the proposed deep learning model achieved an accuracy rate of 95.65%, outperforming traditional models such as ResNet. Beyond precise pollen classification, this technique has broader environmental applications, enabling accurate monitoring of pollen dispersal and contributing to allergy forecasting, biodiversity analysis, climate modeling, and public health initiatives related to pollen exposure. Ultimately, this technology lays the foundation for an integrated air quality monitoring and ecological research system, combining environmental models with advanced monitoring technologies.