Scholarlink

<p>Signal processing presents an inherent contradiction: achieving simultaneous noise suppression and preservation of signal features is exceptionally challenging. To mitigate this issue, this article introduces a novel filtering paradigm termed "Attribute Distance Weighted Average" (ADWA). The fundamental mechanism of this paradigm involves dynamically adjusting the combination of signal attributes through the continuous introduction or replacement of attributes, thereby aiming to mitigate the traditional trade-off between denoising and feature preservation. Within this framework, the selection of attributes significantly influences the filtering outcomes, as the information features associated with different attributes are fundamentally distinct. This study investigates methods to enhance the performance of ADWA in noise suppression by incorporating directional local variance attributes and systematically evaluates the filtering efficacy. The findings indicate that augmenting the filtering dimension through the introduction of key new attributes can enhance the overall performance of the ADWA model, enabling it to excel in both denoising and feature preservation. This study offers specific examples and empirical evidence for the exploration and application of new signal properties within the ADWA framework.</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

The application of directional variance in signal’s denoising based on ADWA

  • 期刊论文

  • 发表日期：

    2026年7月28日

查看全文
  • Gang Xiong,
  • Jisong Zhang,
  • Chenghong Wei,
  • Jun Zhu,
  • Jingye Duan,
  • Zheming Xu

摘要

Signal processing presents an inherent contradiction: achieving simultaneous noise suppression and preservation of signal features is exceptionally challenging. To mitigate this issue, this article introduces a novel filtering paradigm termed "Attribute Distance Weighted Average" (ADWA). The fundamental mechanism of this paradigm involves dynamically adjusting the combination of signal attributes through the continuous introduction or replacement of attributes, thereby aiming to mitigate the traditional trade-off between denoising and feature preservation. Within this framework, the selection of attributes significantly influences the filtering outcomes, as the information features associated with different attributes are fundamentally distinct. This study investigates methods to enhance the performance of ADWA in noise suppression by incorporating directional local variance attributes and systematically evaluates the filtering efficacy. The findings indicate that augmenting the filtering dimension through the introduction of key new attributes can enhance the overall performance of the ADWA model, enabling it to excel in both denoising and feature preservation. This study offers specific examples and empirical evidence for the exploration and application of new signal properties within the ADWA framework.