Scholarlink

This work introduces SparseDeepRain-Net, a novel hybrid architecture meant to improve the removal of rain streaks from photos while maintaining image quality. To successfully distinguish rain streaks from background details, the architecture uses convolutional sparse coding (CSC) and deep image prior (DIP). The suggested model exceeds previous techniques, with a peak signal-to-noise ratio (PSNR) of 37.35 dB and a structural similarity index (SSIM) of 0.95. Furthermore, the model is extremely computationally efficient, with an average processing time of 0.55 s per image. These findings show that SparseDeepRain-Net is a powerful and efficient method for rain removal in image processing applications.

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

Enhancing Image Quality Through Hybrid Sparse Deep Rain-Net: Integrating Convolutional Sparse Coding and Deep Image Prior

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • K. Hemavani,
  • G. S. Annie Grace Vimala,
  • G. NaliniPriya

摘要

This work introduces SparseDeepRain-Net, a novel hybrid architecture meant to improve the removal of rain streaks from photos while maintaining image quality. To successfully distinguish rain streaks from background details, the architecture uses convolutional sparse coding (CSC) and deep image prior (DIP). The suggested model exceeds previous techniques, with a peak signal-to-noise ratio (PSNR) of 37.35 dB and a structural similarity index (SSIM) of 0.95. Furthermore, the model is extremely computationally efficient, with an average processing time of 0.55 s per image. These findings show that SparseDeepRain-Net is a powerful and efficient method for rain removal in image processing applications.