Scholarlink

Abstract <p>The available color space transform (CST) algorithms in image processing pipelines operate suboptimally. The linear (matrix) color coordinate transformation model can theoretically work 1.5–2 times better than the current approach used in modern cameras. Study on the structure of the CST matrix space has shown its low dimensionality: the Hausdorff dimension is about 3 and the effective PCA dimension is about 4. Based on this analysis, new parameterizations for the transformation matrices have been proposed. It has been experimentally confirmed that the proposed parameterizations improve the transform quality by 8.7% as compared with the available methods.</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

Improving the Color Space Transform Quality by Taking into Account the Color Hue of a Light Source

  • 期刊论文

  • 发表日期：

    2025年8月1日

查看全文
  • D. D. Ivanov,
  • V. O. Plokhotnyuk,
  • E. I. Ershov

摘要

Abstract

The available color space transform (CST) algorithms in image processing pipelines operate suboptimally. The linear (matrix) color coordinate transformation model can theoretically work 1.5–2 times better than the current approach used in modern cameras. Study on the structure of the CST matrix space has shown its low dimensionality: the Hausdorff dimension is about 3 and the effective PCA dimension is about 4. Based on this analysis, new parameterizations for the transformation matrices have been proposed. It has been experimentally confirmed that the proposed parameterizations improve the transform quality by 8.7% as compared with the available methods.