Scholarlink

<p>Real-time image processing systems that are highly efficient and optimized are made possible by integrating software and hardware. Noise in images arises from various factors such as low-light conditions, faulty sensors, electronic interference, and transmission errors, all of which compromise image quality and demand effective denoising techniques. This paper proposes a Min-Max Weighted Median filter (MMWMF) for removing impulse noise and an Efficient Median Selector Architecture (EMS) for filtering. Unlike the conventional method, where the weighted median value is substituted for the target pixel value, the proposed approach identifies the location of the median among the weighted values, and the value at that location in the input image is substituted for the target pixel. The proposed median architecture reduces resource utilization by reducing the number of comparisons and is implemented on the xczu7ev-ffvc1156-2-e Zynq UltraScale + ZCU104 Evaluation Board using Vivado software. The results show that the proposed method is more area-, delay-, and power-efficient than the existing median architectures. Compared with the other methods, the proposed EMS-based MMWMF achieves significant reductions in hardware resources. It lowers LUT usage by up to 21.25%, FFs by up to 55.18%, delay by up to 9.78%, and power consumption by up to 13.24%. The method is implemented in MATLAB and compared with other state-of-the-art methods for qualitative purposes.</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

VLSI Implementation of a Non-linear Noise Filtering Technique Using an Efficient Median Selector Architecture

  • 期刊论文

  • 发表日期：

    2025年11月27日

查看全文
  • Chrishia Christudhas,
  • Annis Fathima A

摘要

Real-time image processing systems that are highly efficient and optimized are made possible by integrating software and hardware. Noise in images arises from various factors such as low-light conditions, faulty sensors, electronic interference, and transmission errors, all of which compromise image quality and demand effective denoising techniques. This paper proposes a Min-Max Weighted Median filter (MMWMF) for removing impulse noise and an Efficient Median Selector Architecture (EMS) for filtering. Unlike the conventional method, where the weighted median value is substituted for the target pixel value, the proposed approach identifies the location of the median among the weighted values, and the value at that location in the input image is substituted for the target pixel. The proposed median architecture reduces resource utilization by reducing the number of comparisons and is implemented on the xczu7ev-ffvc1156-2-e Zynq UltraScale + ZCU104 Evaluation Board using Vivado software. The results show that the proposed method is more area-, delay-, and power-efficient than the existing median architectures. Compared with the other methods, the proposed EMS-based MMWMF achieves significant reductions in hardware resources. It lowers LUT usage by up to 21.25%, FFs by up to 55.18%, delay by up to 9.78%, and power consumption by up to 13.24%. The method is implemented in MATLAB and compared with other state-of-the-art methods for qualitative purposes.