Scholarlink

Near-Infrared (NIR) light is used for visualizing veins in medical procedures like venipuncture and intravenous therapy. However, hair artifacts in NIR images hinder accurate vein detection. This study proposes a novel two-phase algorithm to address this issue. Phase one detects hair structures using the Sobel operator, followed by morphological processing to refine detected regions. Phase two applies advanced inpainting techniques to reconstruct vein areas obscured by hair. Tested on nine NIR images, the algorithm achieved an average Mean Squared Error (MSE) of 6.80326, Peak Signal-to-Noise Ratio (PSNR) of 40.39941, and Structural Similarity Index (SSIM) of 0.93261, demonstrating its effectiveness in removing hair artifacts and preserving vein clarity for improved visualization.

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

Hair Removal Algorithm for Enhanced Vein Visualization in Near-Infrared (NIR) Imaging

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mantha Sai Gopal,
  • Satya Prasanna Mallick,
  • Prasanth Tata,
  • Shubham Makwana,
  • A. Parishkrith,
  • Darshan Gera,
  • Vandana Sharma

摘要

Near-Infrared (NIR) light is used for visualizing veins in medical procedures like venipuncture and intravenous therapy. However, hair artifacts in NIR images hinder accurate vein detection. This study proposes a novel two-phase algorithm to address this issue. Phase one detects hair structures using the Sobel operator, followed by morphological processing to refine detected regions. Phase two applies advanced inpainting techniques to reconstruct vein areas obscured by hair. Tested on nine NIR images, the algorithm achieved an average Mean Squared Error (MSE) of 6.80326, Peak Signal-to-Noise Ratio (PSNR) of 40.39941, and Structural Similarity Index (SSIM) of 0.93261, demonstrating its effectiveness in removing hair artifacts and preserving vein clarity for improved visualization.