Automating image segmentation verification and validation by learning test oracles

An image segmentation algorithm delineates (an) object(s) of interest in an image. Its output is referred to as a segmentation. Developing these algorithms is a manual, iterative process involving repetitive verification and validation tasks. This process is time-consuming and depends on the availability of experts, who may be a scarce resource (e.g., medical experts). We propose a framework referred to as Image Segmentation Automated Oracle (ISAO) that uses machine learning to construct an oracle, which can then be used to automatically verify the correctness of image segmentations, thus saving substantial resources and making the image segmentation verification and validation task significantly more efficient. The framework also gives informative feedback to the developer as the segmentation algorithm evolves and provides a systematic means of testing different parametric configurations of the algorithm. During the initial learning phase, segmentations from the first few (optimally two) versions of the segmentation algorithm are manually verified by experts. The similarity of successive segmentations of the same images is also measured in various ways. This information is then fed to a machine learning algorithm to construct a classifier that distinguishes between consistent and inconsistent segmentation pairs (as determined by an expert) based on the values of the similarity measures associated with each segmentation pair. Once the accuracy of the classifier is deemed satisfactory to support a consistency determination, the classifier is then used to determine whether the segmentations that are produced by subsequent versions of the algorithm under test, are (in)consistent with already verified segmentations from previous versions. This information is then used to automatically draw conclusions about the correctness of the segmentations. We have successfully applied this approach to 3D segmentations of the cardiac left ventricle obtained from CT scans and have obtained promising results (accuracies of 95%). Even though more experiments are needed to quantify the effectiveness of the approach in real-world applications, ISAO shows promise in increasing the quality and testing efficiency of image segmentation algorithms.     

Segmentation of a seismic section using image processing and artificial intelligence techniques

Methods for segmenting stacked seismic data into zones of common signal character based on texture analysis are described. Their performance is demonstrated on a line of seismic data from the Gulf of Mexico that had been manually segmented. Two segmentation methods are described. The first is a template matching scheme that matches previously selected data templates with a block of pixels. The second uses statistics determined by examining the run-length of seismic reflection events. The run-length method is extended, through a decision process called the RESOLVER, to incorporate heuristic rules to influence the segmentation. A comparison is made between the automatic segmentations of the section and a manual interpretation.     

Detection of pre-defined boundaries between hydrothermal alteration zones using rotation-variant template matching

A new method for the detection of pre-defined boundaries in single-band image data that uses a rotation-variant template matching (RTM) algorithm is presented. This algorithm matches a miniature image of a pre-defined boundary to image data at various orientations. The image pixels that match boundary criteria are reported in output imagery together with the rotation angle of the template. The method is applied to identify boundaries between hydrothermal alteration zones in processed airborne hyperspectral imagery, based on the presence of white mica minerals. Results show that boundaries identified with RTM are relatively free of noise and more coherent than those identified with, for instance, image slicing techniques. Identified boundaries can be used for image segmentation. The output of the RTM algorithm also provides information on the type of boundary, whether it is crisp or gradual. This information can be used to better characterize mineral variation in the alteration halo associated with fossil hydrothermal systems.     

Segmentation fusion based on neighboring information for MR brain images

In this paper, we study on how to boost image segmentation algorithms. First of all, a novel fusion scheme is proposed to combine different segmentations with mutual information to reduce misclassified pixels and obtain an accurate segmentation. As the class label of each pixel depends on the pixel’s gray level and neighbors’ labels, the fusion scheme takes both spatial and intensity information of pixels into account. Then, a detail thresholding segmentation case is designed using the proposed fusion scheme. In the case, the local Laplacian filter is used to get the smoothed version of original image. To accelerate segmentation, a discrete curve evolution based Otsu method is employed to segment the original image and its smoothed version to get two different segmentation maps. The fusion scheme is used to fuse the two maps to get the final segmentation result. Experiments on medical MR-T2 brain images are conducted to demonstrate the effectiveness of the proposed segmentation fusion method. The experimental results indicate that the proposed algorithm can improve segmentation accuracy and it is superior to other multilevel thresholding methods.     

基于SLIC的自适应多主体图像分割算法

为了解决多主体图像分割的交互分割问题,提出了一种基于SLIC超像素的自适应图像分割算法。首先利用SLIC对图像进行超像素分割处理,把原图像分割为大小相似、形状规则的超像素,以超像素中心点的五维特征值作为原始数据点通过自适应参数的DBSCAN算法聚类,确定多主体数目和分割边界。算法不需要用户交互,自适应确定分割数目。为了验证算法的有效性,在伯克利大学标准数据集BSDS500上与人工标注的分割图像进行比较, 前期的超像素处理使算法在时间上有很好的提升,对于一幅481×321像素的图像,只需要1.5 s就可以获得结果。实验结果表明,该方法可以有效解决多主体图像分割中的人工交互问题,同时在PRI和VOI的指数对比上也优于传统算法,本文算法可以在保证分割效果的基础上自适应确定分割数目,提高分割效率。     

融合T节点线索的图像物体分割

目的 图像分割是计算机视觉、数字图像处理等应用领域首要解决的关键问题。针对现有的单幅图像物体分割算法广泛存在的过分割和过合并现象,提出基于图像T型节点线索的图像物体分割算法。方法 首先,利用L0梯度最小化方法平滑目标图像,剔除细小纹理的干扰;其次,基于Graph-based分割算法对平滑后图像进行适度分割,得到粗糙分割结果;最后,借助于图像中广泛存在的T型节点线索对初始分割块进行区域合并得到最终优化分割结果。结果 将本文算法分别与Grabcut算法及Graph-based算法在不同场景类型下进行了实验与对比。实验结果显示,Grabcut算法需要人工定位边界且一次只能分割单个物体,Graph-based算法综合类内相似度和类间差异性,可以有效保持图像边界,但无法有效控制分割块数量,且分割结果对阈值参数过分依赖,极易导致过分割和过合并现象。本文方法在降低过分割和过合并现象、边界定位精确性和分割准确率方面获得明显改进,几组不同类型的图片分割准确率平均值达到91.16%,明显由于其他算法。处理图像尺寸800×600像素的图像平均耗时3.5 s,较之其他算法略有增加。结论 与各种算法对比结果表明,该算法可有效解决过分割和过合并问题,对比实验结果验证了该方法的有效性,能够取得具有一定语义的图像物体分割结果。     

Dense Features for Semi-Dense Stereo Correspondence

We present a new feature based algorithm for stereo correspondence. Most of the previous feature based methods match sparse features like edge pixels, producing only sparse disparity maps. Our algorithm detects and matches dense features between the left and right images of a stereo pair, producing a semi-dense disparity map. Our dense feature is defined with respect to both images of a stereo pair, and it is computed during the stereo matching process, not a preprocessing step. In essence, a dense feature is a connected set of pixels in the left image and a corresponding set of pixels in the right image such that the intensity edges on the boundary of these sets are stronger than their matching error (which is the difference in intensities between corresponding boundary pixels). Our algorithm produces accurate semi-dense disparity maps, leaving featureless regions in the scene unmatched. It is robust, requires little parameter tuning, can handle brightnessdifferences between images, nonlinear errors, and is fast (linear complexity).     

Integrating region growing and edge detection

A method that combines region growing and edge detection for image segmentation is presented. The authors start with a split-and merge algorithm wherein the parameters have been set up so that an over-segmented image results. Region boundaries are then eliminated or modified on the basis of criteria that integrate contrast with boundary smoothness, variation of the image gradient along the boundary, and a criterion that penalizes for the presence of artifacts reflecting the data structure used during segmentation (quadtree in this case). The algorithms were implemented in the C language on a Sun 3/160 workstation running under the Unix operating system. Simple tool images and aerial photographs were used to test the algorithms. The impression of human observers is that the method is very successful on the tool images and less so on the aerial photograph images. It is thought that the success in the tool images is because the objects shown occupy areas of many pixels, making it is easy to select parameters to separate signal information from noise     

Video sequence motion tracking by fuzzification techniques

In this paper a method for moving objects segmentation and tracking from the so-called permanency matrix is introduced. Our motion-based algorithms enable to obtain the shapes of moving objects in video sequences starting from those image pixels where a change in their grey levels is detected between two consecutive frames by means of the permanency values. In the segmentation phase matching between objects along the image sequence is performed by using fuzzy bi-dimensional rectangular regions. The tracking phase performs the association between the various fuzzy regions in all the images through time. Finally, the analysis phase describes motion through a long video sequence. Segmentation, tracking an analysis phases are enhanced through the use of fuzzy logic techniques, which enable to work with the uncertainty of the permanency values due to image noise inherent to computer vision.     

基于覆盖分割和活动轮廓模型的图像边缘提取算法

针对传统算法对边界模糊的图像分割效果不理想,分割结果多毛刺的问题,提出了一种由粗到细的图像边缘提取方法,主要由像素覆盖分割方法和Chan-Vese模型组成。将改进的覆盖分割方法和活动轮廓模型相结合,首先使用原始覆盖分割算法对图像进行分割,利用多方向模糊形态学边缘检测算法提取不同物体之间的边界;然后采用改进的像素覆盖分割方法给边界像素重新分配覆盖值;最后,运用活动轮廓算法进行细化的图像边界提取;分别进行了分割结果的定性比较,抗噪性测试以及提取的边缘对比实验。实验结果表明,该方法对具有模糊边界的图像,提取边缘结果优于其他可比文献中提出的方法。     

一种具有抗噪性的图像分割方法

基于图论的图像分割方法对有噪声污染的图像必须先进行预处理,算法自身不能抑制噪声。针对该问题,提出一种具有抗噪性的图像分割方法。该方法将图谱划分测度作为划分目标与背景的阈值分割准则,采用基于灰度值的权值矩阵代替基于图像像素个数的权值矩阵,描述像素之间的关联,并在图权计算中增加像素点与其邻域的空间相关信息,以提高算法的抗噪性。实验结果表明,使用该方法进行图像分割具有较好的分割效果,抑制噪声能力较强。     

具有敏感区域的华南虎图像配准研究

对数极坐标映射是一种具有刚体不变性的模板匹配算法,以它获得的刚体变换矩阵为初值经LMA(Levenberg Marquardt algorithm)迭代可以实现图像的高精度配准,但对于场景发生剧烈变化的图像对,上述方法无法获得正确的配准参数。为此,提出了一种具有敏感区域的图像配准算法,在LMA迭代中仅考虑落在人工选取的敏感区域中的像素点,通过对陕西林业厅公布的华南虎照片进行的图像配准实验表明了该算法的有效性。     

基于加速健壮特征拟合算法和Chan-Vese模型的超声图像腔室分割方法

针对超声心动周期序列图的腔室自动分割过程中,弱边缘轮廓难以有效提取的问题,提出一种基于加速健壮特征(SURF)拟合算法和Chan-Vese模型的超声图像腔室分割方法。首先对序列中第一帧图像进行人工标记弱边缘轮廓;然后,提取弱边缘轮廓周围的SURF点,建立Delaunay三角网;接着,通过相邻两帧之间的特征点匹配,预测后续帧的弱边缘轮廓;之后,用Chan-Vese模型提取粗糙轮廓;最后采用区域生长算法得到精确的目标轮廓。实验结果表明,该算法能较好地完整提取超声序列图像中含弱边缘的腔室轮廓,并且与专家手动分割结果相近。     

Data-driven hair segmentation with isomorphic manifold inference

Hair segmentation is challenging due to the diverse appearance, irregular region boundary and the influence of complex background. To deal with this problem, we propose a novel data-driven method, named Isomorphic Manifold Inference (IMI). The IMI method assumes the coarse probability map and the binary segmentation map as a couple of isomorphic manifolds and tries to learn hair specific priors from manually labeled training images. For an input image, firstly, the method calculates a coarse probability map. Then it exploits regression techniques to obtain the relationship between the coarse probability map of the test image and those of training images. Finally, this relationship, i.e., a coefficient set, is transferred to the binary segmentation maps and a soft segmentation of the test image will be achieved by a linear combination of those binary maps. Further, we employ this soft segmentation as a shape cue and integrate it with color and texture cues into a unified segmentation framework. A better segmentation is achieved by the Graph Cuts optimization. Extensive experiments are conducted to validate effectiveness of the IMI method, compare contributions of different cues and investigate the generalization of IMI method. The results strongly encourage our method.     

Image segmentation with ratio cut

This paper proposes a new cost function, cut ratio, for segmenting images using graph-based methods. The cut ratio is defined as the ratio of the corresponding sums of two different weights of edges along the cut boundary and models the mean affinity between the segments separated by the boundary per unit boundary length. This new cost function allows the image perimeter to be segmented, guarantees that the segments produced by bipartitioning are connected, and does not introduce a size, shape, smoothness, or boundary-length bias. The latter allows it to produce segmentations where boundaries are aligned with image edges. Furthermore, the cut-ratio cost function allows efficient iterated region-based segmentation as well as pixel-based segmentation. These properties may be useful for some image-segmentation applications. While the problem of finding a minimum ratio cut in an arbitrary graph is NP-hard, one can find a minimum ratio cut in the connected planar graphs that arise during image segmentation in polynomial time. While the cut ratio, alone, is not sufficient as a baseline method for image segmentation, it forms a good basis for an extended method of image segmentation when combined with a small number of standard techniques. We present an implemented algorithm for finding a minimum ratio cut, prove its correctness, discuss its application to image segmentation, and present the results of segmenting a number of medical and natural images using our techniques.     

A non-stationary MRF model for image segmentation from a soft boundary map

In this paper, we address the problem of estimating a segmentation map into regions from a soft (or possibly probabilistic) boundary representation. For this purpose, we have defined a simple non-stationary MRF model with long-range pairwise interactions whose potentials are estimated from the likelihood of the presence of an edge at each considered pair of pixels. Another contribution of this paper is also to demonstrate that an efficient and interesting alternative strategy to complex and elaborate region-based segmentation models consists in averaging several (quickly estimated) soft contour maps (obtained, for example, with simpler edge-based segmentation models) and to use this MRF reconstruction model to achieve a reliable and accurate segmentation map into regions. This model has been successfully applied on the Berkeley image database. The experiments reported in this paper demonstrate that the proposed segmentation model from an edge map is reliable and that this segmentation strategy is efficient (in terms of visual evaluation and quantitative performance measures) and performs well compared with the best existing state-of-the-art segmentation methods recently proposed in the literature.     

加移动因子的C-V模型

在变分水平集方法中,C-V模型的优点之一是能够提取以非梯度形式定义的图像边界,然而,在提取该类型边界时,模型仅考虑了图像各区域的均值信息而没有考虑图像的局部信息,因此尽管C-V模型能够得到渐进型边界图像的分割结果,但是存在分割误差。将移动因子引入到C-V模型以解决上述问题。其中移动因子定义为图像局部凸凹性的函数,通过该因子可以调整模型0-水平面的高度,进而使得解平面与目标所在平面更加接近或重合,以达到消除分割误差的目的。文中给出了偏微分形式的模型,并通过实验验证了模型的有效性。     

基于形态学重建的粘连物体分割

提出一种基于形态学重建（Morphological Reconstruction）的图像分割方法。该方法先对待分割图像进行预处理,使边界点具有局部极大的灰度值;然后利用灰度形态学重建提取穹顶（Dome）,并根据其特性利用阚值对穹顶进行二值化获得候选边界点集;再利用二值形态学重建确定候选边界点集中的边界点,得到分割边界。实验结果表明,本分割方法所得边界连续性好、假边界少;该方法受噪声和对象内部灰度变化的影响较小,适合用于分割含有粘连对象的图像。     

基于mean-shift全局立体匹配方法

针对图像全局立体匹配精度高、计算量大的问题,提出基于mean shift图像分割的全局立体匹配方法。首先,通过mean shift算法对图像进行分割,获取图像同质区域数量和区域的标号。在计算匹配代价时,根据像素所属的分割区域,对像素进行筛选,从而提高匹配代价计算速度;其次,在代价聚合前,将mean shift算法获取的同质区域数K值赋值给K-means聚类算法,对像素再次聚类,提高立体匹配精度和速度;最后通过TRW-S置信传播解决能量最小化问题。实验表明,该算法明显提高了匹配的准确性和速度,与单纯的全局匹配算法相比,具有更大的优势。     

基于区域相似特征的憎水性图像分割算法

绝缘子憎水性能检测是判断绝缘子抗污闪性能的重要手段。传统的分割算法容易产生误分割而影响判别。为此提出了一种基于目标区域相似性的图像分割算法,根据水珠（水迹)内部具有灰度相似性和梯度相似性的原理进行区域增长,并且运用数学形态学进行图像修正。实验结果表明,该方法实现了目标区域的有效分割,为后续绝缘子性能的判定提供依据。