On Nearest Neighbors in Non Local Means Denoising
This addresses a computational efficiency and quality issue in image denoising, but it is incremental as it modifies an existing method.
The paper tackles the bias introduced by using few nearest neighbors in Non-Local-Means denoising and proposes a Statistical NN criterion, which outperforms the traditional method by producing higher-quality images with fewer neighbors and lower computational cost for both white and colored noise.
To denoise a reference patch, the Non-Local-Means denoising filter processes a set of neighbor patches. Few Nearest Neighbors (NN) are used to limit the computational burden of the algorithm. Here here we show analytically that the NN approach introduces a bias in the denoised patch, and we propose a different neighbors' collection criterion, named Statistical NN (SNN), to alleviate this issue. Our approach outperforms the traditional one in case of both white and colored noise: fewer SNNs generate images of higher quality, at a lower computational cost.