FILTERING ALGORITHMS FOR OPTICAL COHERENCE TOMOGRAPHY IMAGES

The paper provides a comparative analysis of three image filtering algorithms used to improve the quality of optical coherence tomography (OCT): anisotropic diffusion, bilateral filter, and Lee filter. The task of the study was to determine which of the following filters is more suitable for filtering OCT images, which will contribute to more accurate diagnosis of diseases in ophthalmology. Diagnostic data obtained with OCT can be distorted due to the presence of noise, so the correct choice of filtering algorithm is an important step in the image processing process. Anisotropic diffusion is a filtering technique that reduces noise while preserving important image details such as structure boundaries and contours. A bilateral filter is a non-linear filter that takes into account both the spatial arrangement of pixels and their intensity. The Lee filter is an adaptive method that uses local statistical properties of pixels to reduce noise. It works well on images with a uniform texture, providing a high level of noise reduction. The mean square error (MSE) calculation was used to evaluate the effectiveness of the filtering algorithms. MATLAB software was used for image analysis, images with a size of 512x512 pixels are presented in black and white format. The results of the MSE calculations showed that the anisotropic diffusion has smaller values of root mean square error, which indicates its higher efficiency compared to other methods on the given set of images. Thus, the conducted research confirmed that the choice of the optimal filtering algorithm depends on specific tasks and a specific image. The presented filtering algorithms are useful for improving the quality of OCT images, which contributes to increasing the accuracy of ophthalmic diagnostics.