Mohamad Hasan Sayafan

Noise reduction is an important issue in digital image applications. Non-local Means (NL-Means) is an efficient noise reduction where its basis is the employment of self-similarity feature of an image to increase the affectivity of noise reduction. The main problem of this method is its computational cost. This paper proposes an algorithm based on local binary pattern to reduce the computational load of NL-Means method. In this approach, the extra computations are reduced and then, the size of search space, which is used by non-local means, is dramatically declined by PCA. This algorithm remarkably increases the speed of NL-Means and results higher in higher objective quality in terms of PSNR than standard NL-Means.

mage Noise Reduction, Pruning Search Space, Means Algorithm, Binary Pattern Method