Speckle reduction for optical coherence tomography images based on local variance

Speckle reduction has a significant effect on the quality of Optical Coherence Tomography (OCT) images, and anisotropic diffusion is usually used to reduce speckle. In our paper, a method of OCT speckle reduction based on local variance is presented. It obtains a new diffusion coefficient to detect the edge of images more accurate, and the absolute deviation median operator is used to detect the threshold K. In the iteration process of anisotropic diffusion, combinating the discriminate of edges and noises based on local variance further, we process them by two different methods: anisotropy diffusion and mean filter. It can be seen from the final processing results of synthetic images and the OCT image that the proposed algorithm can retain the detail of images better while denoising, In contrast to two other existing speckle reduction algorithms.