Scholarlink

<p>To address the issue of image quality degradation in complex underwater environments, a method combining improved image processing algorithms with deep learning techniques is proposed to enhance underwater image visual quality. Firstly, an improved white balance algorithm is used to correct color cast caused by optical properties, effectively restoring colors. Secondly, logarithmic domain gamma transformation is applied to adjust brightness distribution, enhance dark details, and achieve a more uniform brightness distribution. Additionally, contrast-limited adaptive histogram equalization (CLAHE) is employed to improve image contrast and detail clarity. Finally, bilateral filtering is utilized to remove noise while preserving edge information, which is combined with the U-Net for deblurring. Experimental results demonstrate that the method performs excellently : on the UIEB (Underwater Image Enhancement Benchmark) dataset, the Underwater Image Quality Measure (UIQM) score reaches 2.2758 (a 14.3 % improvement), the Underwater Color Image Quality Evaluation (UCIQE) score is 0.6656 , Peak Signal-to-Noise Ratio (PSNR) is 30.3093dB, and Structural Similarity Index Measure (SSIM) is 0.9623. On the EUVP (Enhancing Underwater Visual Perception) dataset, the UIQM score is 2.2451 (a 12.3 % improvement), UCIQE score is 0.6898 (a 6.6 % improvement), PSNR is 33.4512dB, SSIM is 0.9831 and VIF is 0.9796.</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

Underwater image enhancement methods for complex environments

  • 期刊论文

  • 发表日期：

    2025年5月28日

查看全文
  • Yu Shi,
  • Yue Wang,
  • Hongyu Li

摘要

To address the issue of image quality degradation in complex underwater environments, a method combining improved image processing algorithms with deep learning techniques is proposed to enhance underwater image visual quality. Firstly, an improved white balance algorithm is used to correct color cast caused by optical properties, effectively restoring colors. Secondly, logarithmic domain gamma transformation is applied to adjust brightness distribution, enhance dark details, and achieve a more uniform brightness distribution. Additionally, contrast-limited adaptive histogram equalization (CLAHE) is employed to improve image contrast and detail clarity. Finally, bilateral filtering is utilized to remove noise while preserving edge information, which is combined with the U-Net for deblurring. Experimental results demonstrate that the method performs excellently : on the UIEB (Underwater Image Enhancement Benchmark) dataset, the Underwater Image Quality Measure (UIQM) score reaches 2.2758 (a 14.3 % improvement), the Underwater Color Image Quality Evaluation (UCIQE) score is 0.6656 , Peak Signal-to-Noise Ratio (PSNR) is 30.3093dB, and Structural Similarity Index Measure (SSIM) is 0.9623. On the EUVP (Enhancing Underwater Visual Perception) dataset, the UIQM score is 2.2451 (a 12.3 % improvement), UCIQE score is 0.6898 (a 6.6 % improvement), PSNR is 33.4512dB, SSIM is 0.9831 and VIF is 0.9796.