Robust image segmentation using FCM with spatial constraints based on new kernel-induced distance measure

Fuzzy c-means clustering (FCM) with spatial constraints (FCM_S) is an effective algorithm suitable for image segmentation. Its effectiveness contributes not only to the introduction of fuzziness for belongingness of each pixel but also to exploitation of spatial contextual information. Although the contextual information can raise its insensitivity to noise to some extent, FCM_S still lacks enough robustness to noise and outliers and is not suitable for revealing non-Euclidean structure of the input data due to the use of Euclidean distance (L2 norm). In this paper, to overcome the above problems, we first propose two variants, FCM_S1 and FCM_S2, of FCM_S to aim at simplifying its computation and then extend them, including FCM_S, to corresponding robust kernelized versions KFCM_S, KFCM_S1 and KFCM_S2 by the kernel methods. Our main motives of using the kernel methods consist in: inducing a class of robust non-Euclidean distance measures for the original data space to derive new objective functions and thus clustering the non-Euclidean structures in data; enhancing robustness of the original clustering algorithms to noise and outliers, and still retaining computational simplicity. The experiments on the artificial and real-world datasets show that our proposed algorithms, especially with spatial constraints, are more effective.     

Robust interactive image segmentation using structure-aware labeling

Interactive image segmentation has remained an active research topic in image processing and graphics, since the user intention can be incorporated to enhance the performance. It can be employed to mobile devices which now allow user interaction as an input, enabling various applications. Most interactive segmentation methods assume that the initial labels are correctly and carefully assigned to some parts of regions to segment. Inaccurate labels, such as foreground labels in background regions for example, lead to incorrect segments, even by a small number of inaccurate labels, which is not appropriate for practical usage such as mobile application. In this paper, we present an interactive segmentation method that is robust to inaccurate initial labels (scribbles). To address this problem, we propose a structure-aware labeling method using occurrence and co-occurrence probability (OCP) of color values for each initial label in a unified framework. Occurrence probability captures a global distribution of all color values within each label, while co-occurrence one encodes a local distribution of color values around the label. We show that nonlocal regularization together with the OCP enables robust image segmentation to inaccurately assigned labels and alleviates a small-cut problem. We analyze theoretic relations of our approach to other segmentation methods. Intensive experiments with synthetic and manual labels show that our approach outperforms the state of the art.     

Robust fuzzy clustering-based image segmentation

The fuzzy clustering algorithm fuzzy c-means (FCM) is often used for image segmentation. When noisy image segmentation is required, FCM should be modified such that it can be less sensitive to noise in an image. In this correspondence, a robust fuzzy clustering-based segmentation method for noisy images is developed. The contribution of the study here is twofold: (1) we derive a robust modified FCM in the sense of a novel objective function. The proposed modified FCM here is proved to be equivalent to the modified FCM given by Hoppner and Klawonn [F. Hoppner, F. Klawonn, Improved fuzzy partitions for fuzzy regression models, Int. J. Approx. Reason. 32 (2) (2003) 85–102]. (2) We explore the very applicability of the proposed modified FCM for noisy image segmentation. Our experimental results indicate that the proposed modified FCM here is very suitable for noisy image segmentation.     

Robust color image segmentation using fuzzy c-means with weighted hue and intensity

In this paper we propose a novel method for color image segmentation. This method uses only hue and intensity components (which are chosen rationally) of image and combines those by adaptive tuned weights in a specially defined fuzzy c-means cost function. The tuned weights indicate how informative every color component (hue and intensity) is. Obtaining tuned weights begins with finding peaks of hue and intensity's histograms and continues by obtaining the table of the frequencies of hue and intensity values and computing entropy and contrast of every color component. Also this method specifies proper initial values for cluster centers with the aim of reducing the overall number of iterations and avoiding converging of FCM to wrong centroids. Experimental results demonstrate that our algorithm achieves better segmentation performance and also runs faster than similar methods.     

重要性图约束的均值漂移图像分割

针对已有算法结果分割区域过多问题,提出采用边缘正交场构造重要性图,通过边缘特征稳定性约束分割区域,从而有效地提高分割质量。构造边缘正交场,通过高斯积分提高边缘线的连续性和稳定性。采用边缘特征进行距离变换,生成图像的重要性图。采用均值漂移进行图像预分割,根据相邻区域边界上的重要性强度对分割区域结果进行合并。实验结果表明,和原有分割方法相比较,算法在保持原始图像重要区域的同时,对细节区域进行有效合并,明显提高分割质量。     

Robust estimation for range image segmentation and reconstruction

This correspondence presents a segmentation and fitting method using a new robust estimation technique. We present a robust estimation method with high breakdown point which can tolerate more than 80% of outliers. The method randomly samples appropriate range image points in the current processing region and solves equations determined by these points for parameters of selected primitive type. From K samples, we choose one set of sample points that determines a best-fit equation for the largest homogeneous surface patch in the region. This choice is made by measuring a residual consensus (RESC), using a compressed histogram method which is effective at various noise levels. After we get the best-fit surface parameters, the surface patch can be segmented from the region and the process is repeated until no pixel left. The method segments the range image into planar and quadratic surfaces. The RESC method is a substantial improvement over the least median squares method by using histogram approach to inferring residual consensus. A genetic algorithm is also incorporated to accelerate the random search     

Robust image segmentation against complex color distribution

Color distribution is the most effective cue that is widely adopted in previous interactive image segmentation methods. However, it also may introduce additional errors in some situations, for example, when the foreground and background have similar colors. To address this problem, this paper proposes a novel method to learn the segmentation likelihoods. The proposed method is designed for high reliability, for which purpose it may choose to discard some unreliable likelihoods that may cause segmentation error. The reliability of likelihoods is estimated in a few Expectation–Maximization iterations. In each iteration, a novel multi-class transductive learning algorithm, namely, the Constrained Mapping, is proposed to learn likelihoods and identify unreliable likelihoods simultaneously. The resulting likelihoods then can be used as the input of any segmentation methods to improve their robustness. Experiments show that the proposed method is an effective way to improve both segmentation quality and efficiency, especially when the input image has complex color distribution.     

Robust interactive image segmentation via graph-based manifold ranking

Computational Visual Media - Interactive image segmentation aims at classifying the image pixels into foreground and background classes given some foreground and background markers. In this paper,...     

Exposing image resampling forgery by using linear parametric model

Resampling forgery generally refers to as the technique that utilizes interpolation algorithm to maliciously geometrically transform a digital image or a portion of an image. This paper investigates the problem of image resampling detection based on the linear parametric model. First, we expose the periodic artifact of one-dimensional 1-D) resampled signal. After dealing with the nuisance parameters, together with Bayes’ rule, the detector is designed based on the probability of residual noise extracted from resampled signal using linear parametric model. Subsequently, we mainly study the characteristic of a resampled image. Meanwhile, it is proposed to estimate the probability of pixels’ noise and establish a practical Likelihood Ratio Test (LRT). Comparison with the state-of-the-art tests, numerical experiments show the relevance of our proposed algorithm with detecting uncompressed/compressed resampled images.     

Graphical Gaussian shape models and their application to image segmentation

This paper presents a novel approach to shape modeling and a model-based image segmentation procedure tailor-made for the proposed shape model. A common way to represent shape is based on so-called key points and leads to shape variables, which are invariant with respect to similarity transformations. We propose a graphical shape model, which relies on a certain conditional independence structure among the shape variables. Most often, it is sufficient to use a sparse underlying graph reflecting both nearby and long-distance key point interactions. Graphical shape models allow for specific shape modeling, since, e.g., for the subclass of decomposable graphical Gaussian models both model selection procedures and explicit parameter estimates are available. A further prerequisite to a successful application of graphical shape models in image analysis is provided by the "toolbox" of Markov chain Monte Carlo methods offering highly flexible and effective methods for the exploration of a specified distribution. For Bayesian image segmentation based on a graphical Gaussian shape model, we suggest applying a hybrid approach composed of the well-known Gibbs sampler and the more recent slice sampler. Shape modeling as well as image analysis are demonstrated for the segmentation of vertebrae from two-dimensional slices of computer tomography images.     

Robust skin segmentation using color space switching

Skin detection is very popular and has vast applications among researchers in computer vision and human computer interaction. The skin-color changes beyond comparable limits with considerable change in the nature of the light source. Different properties are taken into account when the colors are represented in different color spaces. However, a unique color space has not been found yet to adjust the needs of all illumination changes that can occur to practically similar objects. Therefore a dynamic skin color model must be constructed for robust skin pixel detection, which can cope with natural changes in illumination. This paper purposes that skin detection in a digital color image can be significantly improved by employing automated color space switching. A system with three robust algorithms has been built based on different color spaces towards automatic skin classification in a 2D image. These algorithms are based on the statistical mean of value of the skin pixels in the image. We also take Bayesian approaches to discriminate between skin-alike and non-skin pixels to avoid noise. This work is tested on a set of images which was captured in varying light conditions from highly illuminated to almost dark.     

Watershed segmentation using prior shape and appearance knowledge

Watershed transformation is a common technique for image segmentation. However, its use for automatic medical image segmentation has been limited particularly due to oversegmentation and sensitivity to noise. Employing prior shape knowledge has demonstrated robust improvements to medical image segmentation algorithms. We propose a novel method for enhancing watershed segmentation by utilizing prior shape and appearance knowledge. Our method iteratively aligns a shape histogram with the result of an improved k-means clustering algorithm of the watershed segments. Quantitative validation of magnetic resonance imaging segmentation results supports the robust nature of our method.     

Color image segmentation using competitive learning

Presents a color image segmentation method which divides the color space into clusters. Competitive learning is used as a tool for clustering the color space based on the least sum-of-squares criterion. We show that competitive learning converges to approximate the optimum solution based on this criterion, theoretically and experimentally. We apply this method to various color scenes and show its efficiency as a color image segmentation method. We also show the effects of using different color coordinates to be clustered, with some experimental results     

Multispectral image compression using eigenregion-based segmentation

In the study, a novel segmentation technique is proposed for multispectral satellite image compression. A segmentation decision rule composed of the principal eigenvectors of the image correlation matrix is derived to determine the similarity of image characteristics of two image blocks. Based on the decision rule, we develop an eigenregion-based segmentation technique. The proposed segmentation technique can divide the original image into some proper eigenregions according to their local terrain characteristics. To achieve better compression efficiency, each eigenregion image is then compressed by an efficient compression algorithm eigenregion-based eigensubspace transform (ER-EST). The ER-EST contains 1D eigensubspace transform (EST) and 2D-DCT to decorrelate the data in spectral and spatial domains. Before performing EST, the dimension of transformation matrix of EST is estimated by an information criterion. In this way, the eigenregion image may be approximated by a lower-dimensional components in the eigensubspace. Simulation tests performed on SPOT and Landsat TM images have demonstrated that the proposed compression scheme is suitable for multispectral satellite image.     

分水岭变换在岩屑图像分割中的应用

分水岭变换是一种基于区域和数学形态学的图像分割方法,被广泛用于灰度图像的分割之中.但传统分水岭变换过分割问题严重,图像的噪声和虚假纹理会淹没真正想得到的边缘信息.针对岩屑图像的特征,提出了一种改进的分水岭算法分割方案.先在预处理期用形态学开闭重建运算对原始图像平滑处理,在相对保留边缘不受影响的同时,降低噪声的影响.再通过非线性的阈值变换分离出目标和背景,然后在提取出目标的情况下合并过小区域,得到目标的边缘.而由于阈值变换后,区域数量已经明显减少,可以降低区域合并的运算量,提高合并速度.在求取形态学梯度时,选用了一种新的形态梯度形式,消除了形态学处理对分割结果造成的轮廓偏移现象.从实验结果看来,该算法取得了较好的分割效果.     

Topology-based image segmentation using LBP pyramids

In this paper, we present a new image segmentation algorithm which is based on local binary patterns (LBPs) and the combinatorial pyramid and which preserves structural correctness and image topology. For this purpose, we define a codification of LBPs using graph pyramids. Since the LBP code characterizes the topological category (local max, min, slope, saddle) of the gray level landscape around the center region, we use it to obtain a “minimal” image representation in terms of the topological characterization of a given 2D grayscale image. Based on this idea, we further describe our hierarchical texture aware image segmentation algorithm and compare its segmentation output and the “minimal” image representation.     

Interactive image segmentation using probabilistic hypergraphs

This paper introduces a novel interactive framework for segmenting images using probabilistic hypergraphs which model the spatial and appearance relations among image pixels. The probabilistic hypergraph provides us a means to pose image segmentation as a machine learning problem. In particular, we assume that a small set of pixels, which are referred to as seed pixels, are labeled as the object and background. The seed pixels are used to estimate the labels of the unlabeled pixels by learning on a hypergraph via minimizing a quadratic smoothness term formed by a hypergraph Laplacian matrix subject to the known label constraints. We derive a natural probabilistic interpretation of this smoothness term, and provide a detailed discussion on the relation of our method to other hypergraph and graph based learning methods. We also present a front-to-end image segmentation system based on the proposed method, which is shown to achieve promising quantitative and qualitative results on the commonly used GrabCut dataset.     

Brain image segmentation using semi-supervised clustering

The objective of brain image segmentation is to partition the brain images into different non-overlapping homogeneous regions representing the different anatomical structures. Magnetic resonance brain image segmentation has large number of applications in diagnosis of neurological disorders like Alzheimer diseases, Parkinson related syndrome etc. But automatically segmenting the MR brain image is not an easy task. To solve this problem, several unsupervised and supervised based classification techniques have been developed in the literature. But supervised classification techniques are more time consuming and cost-sensitive due to the requirement of sufficient labeled data. In contrast, unsupervised classification techniques work without using any prior information but it suffers from the local trap problems. So, to overcome the problems associated with unsupervised and supervised classification techniques, we have proposed a new semi-supervised clustering technique using the concepts of multiobjective optimization and applied this technique for automatic segmentation of MR brain images in the intensity space. Multiple centers are used to encode a cluster in the form of a string. The proposed clustering technique utilizes intensity values of the brain pixels as the features. Additionally it also assumes that the actual class label information of 10% points of a particular image data set is also known. Three cluster validity indices are utilized as the objective functions, which are simultaneously optimized using AMOSA, a modern multiobjective optimization technique based on the concepts of simulated annealing. First two cluster validity indices are symmetry distance based Sym-index and Euclidean distance based I-index, which are based on unsupervised properties. Last one is a supervised information based cluster validity index, Minkowski Index. The effectiveness of this proposed semi-supervised clustering technique is demonstrated on several simulated MR normal brain images and MR brain images having some multiple sclerosis lesions. The performance of the proposed semi-supervised clustering technique is compared with some other popular image segmentation techniques like Fuzzy C-means, Expectation Maximization and some recent image clustering techniques like multi-objective based MCMOClust technique, and Fuzzy-VGAPS clustering techniques.