Effective Image Denoising Model Using Improved Deep Learning Techniques With Optimization Algorithm

Strategic Insights -

Researchers at the Department of Electrical and Communication Engineering have developed a novel image denoising model, combining Improved Convolutional Neural Network (ICNN) with the Self-Improved Orca Predation Algorithm (SI-OPA), to overcome noise distortions in medical images. The proposed model demonstrates superior noise suppression, achieving 94% accuracy, 0.91 Structural Similarity (SSIM), 45.75% Peak Signal-to-Noise Ratio (PSNR), and 43.89% Standard Deviation (STD) metrics.

Key Takeaways:

  • The proposed model, ICNN-SI-OPA, improves denoising efficacy by 8% compared to traditional image denoising algorithms, reaching 94% accuracy.
  • The model's superior performance is attributed to the Self-Improved Orca Predation Algorithm (SI-OPA), which optimizes ICNN's weights, increasing precision of denoising.
  • Comprehensive evaluations reveal ICNN-SI-OPA's competitive edge in picture detail retention, computing efficiency, and noise suppression, outperforming conventional and deep learning-based methods.
  • Experimental findings confirm the viability and efficiency of the proposed model, achieving 88.77% better noise suppression than state-of-the-art approaches.
  • S. Mythili and co-authors' research has been peer-reviewed and published in the International Journal of Pattern Recognition and Artificial Intelligence.
  • The model has been applied to various medical image types, including MRI, CT, X-ray, and ultrasound images, demonstrating its effectiveness in overcoming noise distortions.

Statistics:

  • 94% accuracy, 0.91 SSIM, 45.75% PSNR, and 43.89% STD metrics achieved by the proposed model.
  • 8% improvement in denoising efficacy compared to traditional image denoising algorithms, reaching 94% accuracy.
  • 88.77% better noise suppression than state-of-the-art approaches.
  • 4% increase in precision of denoising with the Self-Improved Orca Predation Algorithm (SI-OPA).
  • Comparative analysis demonstrates competitive edge in picture detail retention, computing efficiency, and noise suppression.

Sources:

  • An Effective Image Denoising Model Using Improved Deep Learning Techniques With Optimization Algorithm. International Journal of Pattern Recognition and Artificial Intelligence, 2025;39(11).
  • S. Mythili et al. (2025). Department of Electrical and Communication Engineering, Rvs Coll Engn, Dindigul 624005, Tamil Nadu, India.