Scholarlink

<p>Satellite image enhancement is crucial for improving feature visibility and ensuring accurate identification in remote sensing applications. However, these images are often degraded by noise and low contrast, which limit analysis accuracy. This paper presents a novel approach for denoising and enhancing satellite images by integrating a Denoised Convolutional Neural Network (D-CNN) with an Enhanced Walrus Optimization (EWO) algorithm. The EWO algorithm optimizes the D-CNN’s parameters, improving image quality by enhancing noise reduction and contrast. Experimental results on the Sentinel satellite dataset show that the proposed model achieves a Peak Signal to Noise Ratio (PSNR) of 39.8 and a Structural Similarity Index (SSIM) of 0.999, outperforming state-of-the-art methods such as DnCNN (PSNR of 38.5, SSIM of 0.988) and Restormer (PSNR of 38.2, SSIM of 0.985). The proposed model demonstrates superior performance in noise reduction and image clarity, offering a reliable and scalable solution for satellite image enhancement in remote sensing applications.</p> Graphical abstract <p></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

Denoising of satellite images using hybridized convolutional neural network

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • M. Rajalakshmi,
  • K. Annapurani Panaiyappan,
  • S. Tamilselvan,
  • A. Bhuvanesh

摘要

Satellite image enhancement is crucial for improving feature visibility and ensuring accurate identification in remote sensing applications. However, these images are often degraded by noise and low contrast, which limit analysis accuracy. This paper presents a novel approach for denoising and enhancing satellite images by integrating a Denoised Convolutional Neural Network (D-CNN) with an Enhanced Walrus Optimization (EWO) algorithm. The EWO algorithm optimizes the D-CNN’s parameters, improving image quality by enhancing noise reduction and contrast. Experimental results on the Sentinel satellite dataset show that the proposed model achieves a Peak Signal to Noise Ratio (PSNR) of 39.8 and a Structural Similarity Index (SSIM) of 0.999, outperforming state-of-the-art methods such as DnCNN (PSNR of 38.5, SSIM of 0.988) and Restormer (PSNR of 38.2, SSIM of 0.985). The proposed model demonstrates superior performance in noise reduction and image clarity, offering a reliable and scalable solution for satellite image enhancement in remote sensing applications.

Graphical abstract