Trilateral filtering for biomedical images

Abstract

Filtering is a core operation in low level computer vision. It is a preliminary process in many biomedical image processing applications. Bilateral filtering has been applied to smooth biomedical images while preserving the edges. However, to avoid oversmoothing structures of sizes comparable to the image resolutions, a narrow spatial window has to be used. This leads to the necessity of performing more iterations in the filtering process. In this paper, we propose a novel filtering technique namely trilateral filter, which can achieve edge-preserving smoothing with a narrow spatial window in only a few iterations. The experimental results have shown that our novel method provides greater noise reduction than bilateral filtering and smooths biomedical images without over-smoothing ridges and shifting the edge locations, as compared to other noise reduction methods.

A nonlinear and non-iterative noise reduction technique for medical images: concept and methods comparison

“…Together with the geometric and photometric similarities, filtering along the orientation of the ridges and gutters can be achieved, whereas low-pass filter is applied to the homogenous regions. For detailed formulation of the methodology, see our related work [10].…”

A gentle introduction to bilateral filtering and its applications

“…In the domain of medical imagery, Wong et al [2004] improve the structure preservation by explicitly accounting the structure with an additional weight depending on the local shape and orientation of the data. …”

Augmented Vessels for Pre-operative Preparation in Endovascular Treatments

“…According to our experiments, a satisfactory 3D RA vascular segmentation can be produced by a global thresholding after the noisy angiography is smoothed with an edge-preserving filter. In this work, we applied the trilateral filter [4] to the angiography for denoising prior to the segmentation.…”

“…After the centerline points are extracted, we present them together with the 3D surface mesh of the vessel lumen to the user 4 . Then the user has to select three centerline points at each end of the vessel of interest.…”