Investigation of Noise Filtering Methods on Histological Images

Abstract

The article presents the results of studies on the reduction and elimination of noise in histological images using filtering methods such as Unsharp mask, Non-local filter, Total variation filter, Block matching and 3d. Studies have shown that the characteristics of the original binary histological image improved more with Non-local filtering than with Unsharp filtering. For color histological images with the possible presence of artifacts, the BM3D filter should be used because it has good performance compared to the Total Variation method. Software has been created to reduce noise on histological images by applying the proposed filters, which can be used to compare the results visually.

Topics

Identifiers

DOI

10.1109/ewdts59469.2023.10297054

Citations and references