Optimization of Denoising Algorithms

In the field of image processing, the three-dimensional block matching algorithm (BM3D) algorithm combines the relevant characteristics of the spatial domain and the frequency domain, and it is one of the algorithms with the best image denoising effect at present, but its denoising thresholds are always qualitative thresholds for different noisy images, which often results in the loss of details in the process of image processing that requires high accuracy, and the noise of the image is not completely removed. This paper proposes a way to optimize the threshold value based on the sand cat swarm algorithm, through the optimization of the threshold value in BM3D, so that the generated threshold curve is more suitable for the threshold value needed by the actual image. The experimental results show that the optimization algorithm denoises the image with reduced running time and improved image quality compared to the original algorithm denoises the image, so that the peak signal-to-noise ratio of the image at a noise standard deviation of 25 is improved from 19.7256 to 21.7862.