Scholarlink

In the realm of computer vision and image processing, local pattern descriptors play a crucial role in extracting discriminative features for various applications. The Equilibrium Difference Local Binary Pattern (ED-LBP) is a recently proposed method that combines color information with local binary patterns, enhancing its robustness and discriminative power. However, the computational demands of ED-LBP is significantly high due to additional mathematical constraints, especially when processing high-resolution images. In this paper, the implementation of ED-LBP is parallelized to improve computational efficiency and enable real-time processing of large-scale image datasets. Our approach leverages the power of Graphics Processing Units (GPUs), resulting in a significant reduction in feature extraction time without compromising the accuracy of the ED-LBP algorithm.

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

Parallel Equilibrium Difference Local Binary Pattern Using GPU

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Uday Yadav,
  • Abhishek G. Sheelvant,
  • B. Ashwath Rao,
  • N. Gopalakrishna Kini

摘要

In the realm of computer vision and image processing, local pattern descriptors play a crucial role in extracting discriminative features for various applications. The Equilibrium Difference Local Binary Pattern (ED-LBP) is a recently proposed method that combines color information with local binary patterns, enhancing its robustness and discriminative power. However, the computational demands of ED-LBP is significantly high due to additional mathematical constraints, especially when processing high-resolution images. In this paper, the implementation of ED-LBP is parallelized to improve computational efficiency and enable real-time processing of large-scale image datasets. Our approach leverages the power of Graphics Processing Units (GPUs), resulting in a significant reduction in feature extraction time without compromising the accuracy of the ED-LBP algorithm.