Scholarlink

<p>Different variables can contribute to the corruption of medical images, causing them to appear blurred or weakened. Therefore, various researches have been done to improve them. This paper utilizes top-hat and bottom-hat morphological transformations with variable-sized disk structuring elements in opening and closing operations, in contrast to previous research that predominantly employed fixed-size disk structuring elements. Additionally, variable coefficients for these transformations are incorporated into the improved image formula, unlike previous studies that utilized constant coefficients. These proposed modifications enable the algorithm to achieve the best possible image enhancement for each individual image. The optimal values ​​for the size of four disks used in transformations and two coefficients are obtained by using genetic algorithm with multi-objective fitting function. Both visual inspection and analysis using metrics like Entropy, SSIM, NIQE, BRISQE, AMBE, and PSNR demonstrate the presented method’s success in enhancing medical images. This improvement is achieved by preserving information, suppressing noise amplification, and preventing significant brightness increases.</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

Improvement of medical images with multi-objective genetic algorithm and adaptive morphological transformations

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月7日

查看全文
  • Sadeq Moradzadeh,
  • Vahid Mehrdad

摘要

Different variables can contribute to the corruption of medical images, causing them to appear blurred or weakened. Therefore, various researches have been done to improve them. This paper utilizes top-hat and bottom-hat morphological transformations with variable-sized disk structuring elements in opening and closing operations, in contrast to previous research that predominantly employed fixed-size disk structuring elements. Additionally, variable coefficients for these transformations are incorporated into the improved image formula, unlike previous studies that utilized constant coefficients. These proposed modifications enable the algorithm to achieve the best possible image enhancement for each individual image. The optimal values ​​for the size of four disks used in transformations and two coefficients are obtained by using genetic algorithm with multi-objective fitting function. Both visual inspection and analysis using metrics like Entropy, SSIM, NIQE, BRISQE, AMBE, and PSNR demonstrate the presented method’s success in enhancing medical images. This improvement is achieved by preserving information, suppressing noise amplification, and preventing significant brightness increases.