Multi-scale-average-filter-assisted level set segmentation model with local region restoration achievements

Abstract

Segmentation of noisy images having light in the background it is a challenging task for the existing segmentation approaches and methods. In this paper, we suggest a novel variational method for joint restoration and segmentation of noisy images which are having intensity and inhomogeneity in the existence of high contrast light in the background. The proposed model combines statistical local region information of circular regions centered at each pixel with a multi-phase segmentation technique enabling inhomogeneous image restoration. The proposed model is written in the fuzzy set framework and resolved through alternating direction minimization approach of multipliers. Through experiments, we have tested the performance of the suggested approach on diverse types of synthetic and real images in the existence of intensity and in-homogeneity; and evaluate the precision, as well as, the robustness of the suggested model. Furthermore, the outcomes are, then, compared with other state-of-the-art models including two-phase and multi-phase approaches and show that our method has superiority for images in the existence of noise and inhomogeneity. Our empirical evaluation and experiments, using real images, evaluate and assess the efficiency of the suggested model against several other closest rivals. We observed that the suggested model can precisely segment all the images having brightness, diffuse edges, high contrast light in the background, and inhomogeneity.

A Convex Selective Segmentation Model based on a Piece-wise Constant Metric Guided Edge Detector Function

“…Even though, the MS minimization model is non-convex and very challenging. On the other hand, the edge based methods in comparison with region based methods have, as reported in previous studies 12,39,44 , the main drawback which is the lack of robustness in dealing with noisy images. To get the benefits of both region and edge based valuable properties, leading to better segmentation accuracy and robustness, mixed models have been introduced in the literature [5][6][7] .…”

“…Recently, the work presented in 43 included weight function into the MS model and proposed a second-order convex model for image segmentation, in particular, selective segmentation. Even though, the above-mentioned methods show good numerical performance there are still some challenging images demonstrating their fail 44,45 .…”