A Comparison of Affine Region Detectors

The paper gives a snapshot of the state of the art in affine covariant region detectors, and compares their performance on a set of test images under varying imaging conditions. Six types of detectors are included: detectors based on affine normalization around Harris (Mikolajczyk and Schmid, 2002; Schaffalitzky and Zisserman, 2002) and Hessian points (Mikolajczyk and Schmid, 2002), a detector of ‘maximally stable extremal regions', proposed by Matas et al. (2002); an edge-based region detector (Tuytelaars and Van Gool, 1999) and a detector based on intensity extrema (Tuytelaars and Van Gool, 2000), and a detector of ‘salient regions', proposed by Kadir, Zisserman and Brady (2004). The performance is measured against changes in viewpoint, scale, illumination, defocus and image compression. The objective of this paper is also to establish a reference test set of images and performance software, so that future detectors can be evaluated in the same framework. First online version published in October, 2005     

Object Detection using Geometrical Context Feedback

We propose a new coherent framework for joint object detection, 3D layout estimation, and object supporting region segmentation from a single image. Our approach is based on the mutual interactions among three novel modules: (i)?object detector; (ii)?scene 3D layout estimator; (iii)?object supporting region segmenter. The interactions between such modules capture the contextual geometrical relationship between objects, the physical space including these objects, and the observer. An important property of our algorithm is that the object detector module is capable of adaptively changing its confidence in establishing whether a certain region of interest contains an object (or not) as new evidence is gathered about the scene layout. This enables an iterative estimation procedure where the detector becomes more and more accurate as additional evidence about a specific scene becomes available. Extensive quantitative and qualitative experiments are conducted on the table-top dataset (Sun et?al. in ECCV, 2010b) and two publicly available datasets (Hoiem et?al. in CVPR, 2006; Sudderth et?al. in IJCV, 2008), and demonstrate competitive object detection, 3D layout estimation, and segmentation results.     

一种基于Zernike矩的局部特征检测方法

针对图像发生几何或质量畸变时局部特征区域提取效果不理想的问题,提出了一种基于Zernike矩的具有旋转不变性与尺度不变性的图像局部特征检测算子。该算法利用Zernike矩构建Hessian矩阵,以基于Zernike矩的Hessian矩阵的行列式与迹确定潜在兴趣点的位置,使用非极大值抑制获得多尺度模板下的最大角点响应,再经二维二次插值运算精确定位兴趣点位置,最后利用主曲率进行边缘响应抑制,利用梯度方向直方图确定兴趣点主方向,由兴趣点4×4邻域的8个方向构建描述算子。实验结果表明,该特征检测方法在视角变换、旋转缩放、图像模糊、图像压缩以及光照变化等图像畸变条件下是有效的,且具有良好的抗噪性能。     

Artifact analysis on triangular box-splines and subdivision surfaces defined by triangular polyhedra

Surface artifacts are features in a surface which cannot be avoided by movement of control points. They are present in B-splines, box splines and subdivision surfaces. We showed how the subdivision process can be used as a tool to analyse artifacts in surfaces defined by quadrilateral polyhedra ( [Sabin et al., 2005] and [Augsd?rfer et al., 2011]).In this paper we are utilising the subdivision process to develop a generic expression which can be employed to determine the magnitude of artifacts in surfaces defined by any regular triangular polyhedra. We demonstrate the method by analysing box-splines and regular regions of subdivision surfaces based on triangular meshes: Loop subdivision, Butterfly subdivision and a novel interpolating scheme with two smoothing stages. We compare our results for surfaces defined by triangular polyhedra to those for surfaces defined by quadrilateral polyhedra.     

A No Free Lunch theorem for multi-objective optimization

The No Free Lunch theorem (Schumacher et al., 2001; Wolpert and Macready, 1997 [8] and [10]) is a foundational impossibility result in black-box optimization stating that no optimization technique has performance superior to any other over any set of functions closed under permutation.This paper considers situations in which there is some form of structure on the set of objective values other than the typical total ordering (e.g., Pareto dominance in multi-objective optimization). It is shown that in such cases, when attention is restricted to natural measures of performance and optimization algorithms that measure performance and optimize with respect to this structure, that a No Free Lunch result holds for any class of problems which is structurally closed under permutation. This generalizes the Sharpened No Free Lunch theorem of Schumacher et al. (2001) [8] to non-totally ordered objective spaces.     

A hybrid approach to outlier detection based on boundary region

In recent years, much attention has been given to the problem of outlier detection, whose aim is to detect outliers - objects who behave in an unexpected way or have abnormal properties. The identification of outliers is important for many applications such as intrusion detection, credit card fraud, criminal activities in electronic commerce, medical diagnosis and anti-terrorism, etc. In this paper, we propose a hybrid approach to outlier detection, which combines the opinions from boundary-based and distance-based methods for outlier detection ( [Jiang et al., 2005], [Jiang et al., 2009] and [Knorr and Ng, 1998]). We give a novel definition of outliers - BD (boundary and distance)-based outliers, by virtue of the notion of boundary region in rough set theory and the definitions of distance-based outliers. An algorithm to find such outliers is also given. And the effectiveness of our method for outlier detection is demonstrated on two publicly available databases.     

A p-multigrid spectral difference method for two-dimensional unsteady incompressible Navier-Stokes equations

This paper presents the development of a 2D high-order solver with spectral difference method for unsteady incompressible Navier-Stokes equations accelerated by a p-multigrid method. This solver is designed for unstructured quadrilateral elements. Time-marching methods cannot be applied directly to incompressible flows because the governing equations are not hyperbolic. An artificial compressibility method (ACM) is employed in order to treat the inviscid fluxes using the traditional characteristics-based schemes. The viscous fluxes are computed using the averaging approach (Sun et al., 2007; Kopriva, 1998) [29] and [12]. A dual time stepping scheme is implemented to deal with physical time marching. A p-multigrid method is implemented (Liang et al., 2009) [16] in conjunction with the dual time stepping method for convergence acceleration. The incompressible SD (ISD) method added with the ACM (SD-ACM) is able to accurately simulate 2D steady and unsteady viscous flows.     

Class-specific image representation for image classification using multiple scale-invariant region detectors

We propose a new class-specific image representation for image classification using multiple region detectors. The new representation is designed to solve the problem of increasing variation in object location and size within images of a class, for which traditional spatial pyramid matching shows limited classification accuracy. We propose a new region-division method that divides the image region into two class-specific regions, called class-specific region-of-interest (C-ROI) and focal region (FR). Using multiple region detectors and appropriate mixing of their responses avoids the problem of selecting a region detector that gives the best classification accuracy for a given image class, and thereby yields better results than using only one region detector. Several scale-invariant region detectors are used to obtain C-ROI and FR by considering their importance over a given image class. In experiments using several well-known datasets, the proposed method improved the accuracy and achieved results that were better than or comparable to those achieved by the related methods.     

Visualizing references to off-screen content on mobile devices: A comparison of Arrows, Wedge, and Overview + Detail

When navigating large information spaces on mobile devices, the small size of the display often causes relevant content to shift off-screen, greatly increasing the difficulty of spatial tasks such as planning routes or finding points of interest on a map. Two possible approaches to mitigate the problem are Contextual Cues, i.e. visualizing abstract shapes in the border region of the view area to function as visual references to off-screen objects of interest, and Overview + Detail, i.e., simultaneously displaying a detail view and a small-scale overview of the information space. In this paper, we compare the effectiveness of two different Contextual Cues techniques, Wedge (Gustafson et al., 2008) and Scaled Arrows (Burigat et al., 2006), and a classical Overview + Detail visualization that highlights the location of objects of interest in the overview. The study involved different spatial tasks and investigated the scalability of the considered visualizations, testing them with two different numbers of off-screen objects. Results were multifaceted. With simple spatial tasks, no differences emerged among the visualizations. With more complex spatial tasks, Wedge had advantages when the task required to order off-screen objects with respect to their distance from the display window, while Overview + Detail was the best solution when users needed to find those off-screen objects that were closest to each other. Finally, we found that even a small increase in the number of off-screen objects negatively affected user performance in terms of accuracy, especially in the case of Scaled Arrows, while it had a negligible effect in terms of task completion times.     

迭代分割逼近:新的基于局部响应显著度的角点检测

为了提高角点检测算法的精确度,同时保持较低的时间复杂度,提出了一种基于自定义的局部角点响应显著度和迭代分割逼近的角点检测方法。首先,定义了一个新的测度量——局部角点响应显著度(LCRS),用来衡量一个候选角点在其局部区域内的响应显著程度,并证明了基于LCRS的角点检测准则等价于局部自适应阈值法。其次,将LCRS视作区域的响应显著度性质,把寻找角点的任务转化为寻找高显著性区域。据此,提出迭代分割的策略用来逐步收缩显著区域,最终逼近真实角点的位置。迭代分割逼近(ISA)算法可以使用不同的角点响应函数(CRF)定义,而且其平均情况时间复杂度与Harris算法相同。实验结果表明,当采用Noble算子的CRF时,ISA算法平均误检率、漏检率分别比Noble算子低4.62%和5.59%;而当采用Harris算子的CRF时,这两个比率也分别比Harris算子低2.87%和3.37%。而且这两种情况下ISA算法的平均运行时间均小于Harris算子和Noble算子。     

基于Gabor多尺度空间的不变兴趣点检测

针对以往仿射不变兴趣点的特征尺度不能直接断定的问题, 提出一种基于Gabor多尺度空间的不变兴趣点检测算法。该算法主要包括三个步骤：应用Gabor滤波器组与图像卷积建立图像Gabor多尺度空间; 通过极大值准则检测兴趣点并直接断定特征尺度; 采用二阶矩矩阵描述兴趣点局部结构。实验结果表明, 相比较其他Hessian-Affine、MSER等算法, 该算法在图像模糊和JPEG压缩情况下可重复率和可匹配率均取得最好结果, 是一种能有效直接提取特征尺度的兴趣点检测算法。     

From Images to Shape Models for Object Detection

We present an object class detection approach which fully integrates the complementary strengths offered by shape matchers. Like an object detector, it can learn class models directly from images, and can localize novel instances in the presence of intra-class variations, clutter, and scale changes. Like a shape matcher, it finds the boundaries of objects, rather than just their bounding-boxes. This is achieved by a novel technique for learning a shape model of an object class given images of example instances. Furthermore, we also integrate Hough-style voting with a non-rigid point matching algorithm to localize the model in cluttered images. As demonstrated by an extensive evaluation, our method can localize object boundaries accurately and does not need segmented examples for training (only bounding-boxes).     

Silhouette-based occluded object recognition through curvature scale space

A complete and practical system for occluded object recognition has been developed which is very robust with respect to noise and local deformations of shape (due to weak perspective distortion, segmentation errors and non-rigid material) as well as scale, position and orientation changes of the objects. The system has been tested on a wide variety of free-form 3D objects. An industrial application is envisaged where a fixed camera and a light-box are utilized to obtain images. Within the constraints of the system, every rigid 3D object can be modeled by a limited number of classes of 2D contours corresponding to the object's resting positions on the light-box. The contours in each class are related to each other by a 2D similarity transformation. The Curvature Scale Space technique [26, 28] is then used to obtain a novel multi-scale segmentation of the image and the model contours. Object indexing [16, 32, 36] is used to narrow down the search space. An efficient local matching algorithm is utilized to select the best matching models. Received: 5 August 1996 / Accepted: 19 March 1997     

局部不变特征综述

局部不变特征是近年来计算机视觉领域的研究热点。局部不变特征在宽基线匹配、特定目标识别、目标类别识别、图像及视频检索、机器人导航、场景分类、纹理识别和数据挖掘等多个领域得到了广泛的应用。本文基于局部不变特征检测、局部不变特征描述和局部不变特征匹配3个基本问题,综述了文献中现有的局部不变特征研究方法,并比较了各类方法的优缺点。根据特征层次的不同,局部不变特征检测方法可以分为角点不变特征、blob不变特征和区域不变特征检测方法3类。局部不变特征的描述方法可以分为基于分布的描述方法、基于滤波的描述方法、基于矩的描述方法和其他描述方法。局部不变特征匹配的研究主要集中在相似性度量、匹配策略和匹配验证3个方面。最后在分析各类研究方法的基础上,总结了局部不变特征研究目前存在的一些问题及可能的发展方向。