Minimization of Region-Scalable Fitting Energy for Image Segmentation

By Yunyun Yang, Chunming Li, Chiu Yen Kao, Stanley J. Osher. IEEE Transactions on Image Processing, 2008

1,794 citations recorded by the source corpus.

Intensity inhomogeneities often occur in real-world images and may cause considerable difficulties in image segmentation. In order to overcome the difficulties caused by intensity inhomogeneities, we propose a region-based active contour model that draws upon intensity information in local regions at a controllable scale. A data fitting energy is defined in terms of a contour and two fitting functions that locally approximate the image intensities on the two sides of the contour. This energy is then incorporated into a variational level set formulation with a level set regularization term, from which a curve evolution equation is derived for energy minimization. Due to a kernel function in the data fitting term, intensity information in local regions is extracted to guide the motion of the contour, which thereby enables our model to cope with intensity inhomogeneity. In addition, the regularity of the level set function is intrinsically preserved by the level set regularization term to ensure accurate computation and avoids expensive reinitialization of the evolving level set function. Experimental results for synthetic and real images show desirable performances of our method.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/1840d7418696f4c225fad8b7dc57a34d041a9273

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Segmentation of X‐ray computed tomography images of porous materials: A crucial step for characterization and quantitative analysis of pore structures (2009, methodology): “Other examples include methods that combine level set methods with simple clustering [ Li et al., 2008 ] or Bayesian analysis [Zhou et al., 2008] to account for noise and heterogeneity of image intensity.”
TouchCut: Fast image and video segmentation using single-touch interaction (2014, background): “This approach to color consistency, combined with a novel per-pixel adaptive weighting scheme, mitigates the tendency for contour expansion to skip the real boundary when color models of the foreground and background are indistinct.”
A Level Set Method for Image Segmentation in the Presence of Intensity Inhomogeneities With Application to MRI (2011, background): “The energy term is defined by (19) which computes the arc length of the zero level contour of and therefore serves to smooth the contour by penalizing its arc length [4] , [10].”
Deep Level Sets for Salient Object Detection (2017, methodology): “With new kind of methods, additional information like region [8, 27], shape [49, 52] can be conveniently and naturally formulated in the level set domain.”
A survey of MRI-based brain tumor segmentation methods (2014, methodology): “A region-based active contour model was proposed in a variable level set formulation for the segmentation of brain tumor [125] .”
Computer Methods and Programs in Biomedicine (neutral): “To overcome this limit, a localized region energy formulation is introduced in [183] , by replacing E 1 and E 2 with E loc 1 and E loc 2 : E loc being w a weighting function, and f 1 and f 2 functions that approximate the image intensity inside and outside S , respectively.”
Active contours with selective local or global segmentation: A new formulation and level set method (2010, positive): “(5), the level set formulation of the proposed model is as follows: @/ @t ¼ spf IðxÞð Þ div r/jr/j þ a jr/j þ rspf ðIðxÞÞ r/; x 2 X ð13Þ”
Active contours driven by local Gaussian distribution fitting energy (2009, methodology): “The LBF model [25,23,31] was recently proposed to segment images with intensity inhomogeneity, using the local intensity information efficiently, and has achieved promising results.”
Histology image analysis for carcinoma detection and grading (2012, methodology): “…understand these challenges, in this section we review the principles of currently commonly used image segmentation methods, including multithresholding [53], clustering-based methods (K-means [44] and mean shift [40]), and state-of-the-art methods using MRF [11,17,158] and active contours [27,93].”
Wound intensity correction and segmentation with convolutional neural networks (2017, background): “In general, ACMs can be divided into two categories: edge-based [7, 8] and region-based [9, 10].”
Distance Regularized Level Set Evolution and Its Application to Image Segmentation (2010, background): “These desirable features, among others, have greatly promoted further development of level set methods and their applications in image segmentation [10]–[12], [7], [13]–[16], tracking [14], [17], and stereo reconstruction [18], etc.”
State of the art survey on MRI brain tumor segmentation. (2013, background): “…severe problems for post-processing of the data sets and medical diagnostics [11], in the last decade several robust algorithms for correcting intensity inhomogeneities in MRI images have been proposed [12 – 15], making the ability to compensate for differences in illumination less of an advantage.”
A visual analytics approach to understanding cycling behaviour (2012, neutral): “Arcoumanis et al. [3], were one of the first who studied experimentally the breakup of Diesel droplets using the continuous air jet arrangement.”