Effectiveness of Image Curvelet Transform Coefficients for Image Denoising

In this research, we investigate the effect of image curvelet transform coefficients in image denoising. The curvelet transform applies to input images, resulting in a set of curvelet coefficients that capture different frequency and directional components of the image. To improve the denoising process, we introduce an approach (CTRLC: Curvelet Transformation Remove Least Correlation) based on the correlation between the abstract coefficient and other coefficients. By analyzing the correlation values, we identify the coefficient that is least associated with the abstract coefficient and remove it from the transformed image. This selective removal allows us to attenuate noise while preserving the relevant image information. Experimental evaluations are conducted on a variety of images contaminated with different levels of noise. The results show that our proposed method effectively reduces noise and enhances the image quality. Comparative analyses with existing denoising techniques further validate the superiority of our approach in terms of noise reduction and preservation of important image details. The CTRLC method achieved a PSNR of 87.2695, compared to other methods that ranged between 23.43 and 77.5. This confirms the effectiveness of our proposed approach in image restoration after denoising. The findings of this research contribute to the field of image denoising by highlighting the significance of curvelet transform coefficients and the correlation-based coefficient removal technique. The proposed method offers a solution for effectively reducing noise in images while maintaining their visual integrity.