Scholarlink

This paper presents scientific and methodological methods for selecting spectral transformation algorithms for processing muscle activity biosignals. Discrete Cosine Transform, Fast Fourier Transform and Hadamard Transform were selected as the objects. Algorithms that transform signals from the time domain to the frequency domain were compared based on the metrics Energy Reduction, Reconstruction Error, Processing Time, Root Mean Square Error, Correlation, Spectral Distortion, Energy Compaction, and the results are presented. Because of scientific research, Fast Fourier Transform (FFT) showed the best results in terms of processing time—1.44 ms, reconstruction error—6% and energy compaction—99.97%. Discrete Cosine Transform (DCT) showed high results in terms of energy reduction—94% and energy compaction—97.98%, but Hadamard Transform (HT) showed lower results in many indicators.

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

Scientific Foundations for Selecting Spectral Transformation Algorithms in Electromyography Signal Processing

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Kamoliddin Shukurov,
  • Mekhriddin Mirjamolov,
  • A’lokhan Kakhkharov,
  • Shokhrukhmirzo Kholdorov

摘要

This paper presents scientific and methodological methods for selecting spectral transformation algorithms for processing muscle activity biosignals. Discrete Cosine Transform, Fast Fourier Transform and Hadamard Transform were selected as the objects. Algorithms that transform signals from the time domain to the frequency domain were compared based on the metrics Energy Reduction, Reconstruction Error, Processing Time, Root Mean Square Error, Correlation, Spectral Distortion, Energy Compaction, and the results are presented. Because of scientific research, Fast Fourier Transform (FFT) showed the best results in terms of processing time—1.44 ms, reconstruction error—6% and energy compaction—99.97%. Discrete Cosine Transform (DCT) showed high results in terms of energy reduction—94% and energy compaction—97.98%, but Hadamard Transform (HT) showed lower results in many indicators.