Scholarlink

<p>To address the challenge of color reproduction in yarns produced by three-channel CNC spinning, this study proposes a novel color prediction approach based on the Friele algorithm and a cylindrical model constructed from nine primary colors. Within this model, 204 grid point mixed samples served as training data, while 12 non-grid point mixed samples were used for testing samples. The core innovation lies in the algorithm’s parameterization strategy: model parameters were established for three primary color samples and one mixed sample, with optimal parameters (minimizing color difference) identified through iterative refinement. Crucially, the method implements a differentiated prediction scheme: reflectance for grid point samples is predicted directly using their model parameters, whereas reflectance for non-grid point samples is determined indirectly using model parameters from adjacent grid point mixed samples. Experimental validation demonstrated promising results: the testing samples exhibited a maximum color difference of 4.88, a minimum of 1.02, and an average of 2.52. These findings confirm the significant potential of this method for accurate color prediction of mélange yarns manufactured via three-channel CNC spinning.</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

Research on Friele Algorithm from the Full Color Gamut Mixing Model

  • 期刊论文

  • 发表日期：

    2026年4月23日

查看全文
  • Xianqiang Sun,
  • Weiqiang Xu,
  • Yuan Xue

摘要

To address the challenge of color reproduction in yarns produced by three-channel CNC spinning, this study proposes a novel color prediction approach based on the Friele algorithm and a cylindrical model constructed from nine primary colors. Within this model, 204 grid point mixed samples served as training data, while 12 non-grid point mixed samples were used for testing samples. The core innovation lies in the algorithm’s parameterization strategy: model parameters were established for three primary color samples and one mixed sample, with optimal parameters (minimizing color difference) identified through iterative refinement. Crucially, the method implements a differentiated prediction scheme: reflectance for grid point samples is predicted directly using their model parameters, whereas reflectance for non-grid point samples is determined indirectly using model parameters from adjacent grid point mixed samples. Experimental validation demonstrated promising results: the testing samples exhibited a maximum color difference of 4.88, a minimum of 1.02, and an average of 2.52. These findings confirm the significant potential of this method for accurate color prediction of mélange yarns manufactured via three-channel CNC spinning.