Unsupervised deformable image registration network for 3D medical images

Image registration aims to establish an active correspondence between a pair of images. Such correspondence is critical for many significant applications, such as image fusion, tumor growth monitoring, and atlas generation. In this study, we propose an unsupervised deformable image registration network (UDIR-Net) for 3D medical images. The proposed UDIR-Net is designed in an encoder-decoder architecture and directly estimates the complex deformation field between input pairwise images without any supervised information. In particular, we recalibrate the feature slice of each feature map that is propagated between the encoder and the decoder in accordance with the importance of each feature slice and the correlation between feature slices. This method enhances the representational power of feature maps. To achieve efficient and robust training, we design a novel hierarchical loss function that evaluates multiscale similarity loss between registered image pairs. The proposed UDIR-Net is tested on different public magnetic resonance image datasets of the human brain. Experimental results show that UDIR-Net exhibits competitive performance against several state-of-the-art methods.

     

视觉文档图像的几何校正

在以数码相机等数字设备拍摄文档资料时,所拍摄的图像经常会产生各种各样的几何变形,这种变形可能会导致识别软件中的版面分析和切分算法失效,从而使文档图像无法被识别。为了使普通的识别软件能够对数码相机等拍摄的文档图像进行识别,有必要对其进行几何校正。根据几何变形产生的原因对其进行了分类,并针对不同种类的变形提出了相应的校正算法。实验结果证明该分类方法和相应校正算法都有较好的效果。     

一种三维CT图像中的线状目标检测方法

当前三维CT图像广泛应用在工业和医学中, 对于工业无损检测和医学上的病灶分析具有重要的研究意义。线状目标广泛存在于医学和工业的CT图像中, 为了实现三维空间中线状目标检测, 传统的直线检测方法有Hough和Radon变换, 但是计算量很大, 而且不适合曲线检测, 对于三维图像来说, 计算更为复杂。因此提出一种基于距离变换, 并通过端点和拐点检测提取线状目标检测算法, 不仅对于直线目标有较好的检测效果, 对特定曲线也有较好的检测结果, 而且通过检测距离变换的距离, 自动地检测线状目标的粗细尺度属性。实验证明, 该方法具有较好的检测结果。     

立体图像对的极线校正

将Seitz方法和Francesco方法结合起来,给出了一种改进的极线校正方法.该方法利用改进的八点算法算出基本矩阵,根据基本矩阵采用Seitz方法计算投影变换矩阵,最后以算出的投影变换矩阵为初值采用Francesco方法对其进行最优化,从而有效地避免了最优化时陷入局部最优的可能,同时又不过分依赖基本矩阵的精度.通过对Francesco方法、Mallon方法和该方法进行对比实验,可以发现该方法具有更高的校正精度,图像的扭曲也比较小.     

基于单幅折反射全向图的水平直线3维重建

与传统的利用立体视觉原理进行3维重建不同,在研究基于单幅全向图像重建空间水平直线问题的基础上,首先指出基于直线在全向图中的两个像点即可重建该水平直线,然后通过分析和推导空间水平直线在全向成像系统中的成像特点,有效简化了全向图中的水平直线检测;针对现有“四点定位”方法的不足,提出了一种基于“主像点/非主像点”的水平直线重建算法,并详细分析了像点提取精度对直线重建结果的影响。实验表明,在不同的像点提取精度下,对于不同空间水平直线,该重建算法均能取得较好结果。     

Numerical processing of CNT arrays using 3D image processing of SEM images

This paper outlines an image processing based technique to characterize carbon nanotube (CNT) array devices for enhanced cell transfection. Investigating how manufacturing parameters affects CNT array geometry, and how geometry affects transfection, requires the arrays to be measured. Obtaining a statistically sufficient number of measurements by hand is tedious and subject to human error. An automated system to characterize the arrays facilitates data collection of numerous pore properties. Scanning electron microscopy (SEM) micrographs are pre-processed to identify the location of CNTs which are then measured individually to obtain their characteristics. The data from single pores is aggregated to generate a numerical summary of the array parameters. Stereomicroscopy techniques are used to measure the heights of the CNTs using pairs of tilted images. The overall technique accurately measures the parameters relevant to cell transfection significantly faster than manual measurements while eliminating human error and bias.     

Winding and Euler numbers for 2D and 3D digital images

New algorithms for computing the Euler number of a 3D digital image S are given, based on smoothing the image to a differentiable object and applying theorems of differential geometry and algebraic topology. They run in O(n) time, where n is the number of object elements of S with neighbors not in S. The basic idea is general and easily extended to images defined by other means, such as a hierarchical data structure or a union of isothetic (hyper) rectangles.     

Metric rectification of curved document images

In this paper, we propose a metric rectification method to restore an image from a single camera-captured document image. The core idea is to construct an isometric image mesh by exploiting the geometry of page surface and camera. Our method uses a general cylindrical surface (GCS) to model the curved page shape. Under a few proper assumptions, the printed horizontal text lines are shown to be line convergent symmetric. This property is then used to constrain the estimation of various model parameters under perspective projection. We also introduce a paraperspective projection to approximate the nonlinear perspective projection. A set of close-form formulas is thus derived for the estimate of GCS directrix and document aspect ratio. Our method provides a straightforward framework for image metric rectification. It is insensitive to camera positions, viewing angles, and the shapes of document pages. To evaluate the proposed method, we implemented comprehensive experiments on both synthetic and real-captured images. The results demonstrate the efficiency of our method. We also carried out a comparative experiment on the public CBDAR2007 data set. The experimental results show that our method outperforms the state-of-the-art methods in terms of OCR accuracy and rectification errors.     

Quasi-Euclidean epipolar rectification of uncalibrated images

This paper deals with the problem of epipolar rectification in the uncalibrated case. First the calibrated (Euclidean) case is recognized as the ideal one, then we observe that in that case images are transformed with a collineation induced by the plane at infinity, which has a special structure. Hence, that structure is imposed to the sought transformation while minimizing a rectification error. Experiments show that this method yields images that are close to the ones produced by Euclidean rectification.     

海面透视图像的度量矫正方法

在海洋测绘中,通过图像矫正的方法估算出测量仪器的位置具有非常重要的意义。针对海面目标位置估算中的图像透视矫正问题,提出一种基于平面虚圆点图像的海面图像度量矫正方法。利用不同角度下照片中圆形特征物的图像计算出绝对二次曲线的图像;根据绝对二次曲线以及图像中的海平线辨认海平面虚圆点的图像;通过对与海平面虚圆点对偶的二次曲线的图像进行奇异值分解,获得海平面图像的度量矫正矩阵。在图像实验中,平面上的角度矫正误差小于1.3°,表明这种方法具有较好的矫正精度,能够满足多种海洋测绘作业的要求。     

Reconstruction of 3D medical images: A nonlinear interpolation technique for reconstruction of 3D medical images

Three-dimensional medical images reconstructed from a series of two-dimensional images produced by computerized tomography, magnetic resonance imaging, etc., present a valuable tool for modern medicine. Usually, the interresolution between two cross sections is less than the intraresolution within each cross section. Therefore, interpolations are required to create a 3D visualization. Many techniques, including voxel-based and patch tiling methods, apply linear interpolations between two cross sections. Although those techniques using linear interpolations are economical in computation, they need much cross-sectional data and are unable to enlarge because of aliasing. Hence, the techniques that apply two-dimensional nonlinear interpolation functions among cross sections were proposed. In this paper, we introduce the curvature sampling of the contour of a medical object in a CT (computerized tomography) image. Those sampled contour points are the candidates for the control points of Hermite surfaces between each pair of cross sections. Then, a nearest-neighbor mapping of control points between every two cross sections is used for surface formation. The time complexity of our mapping algorithm is O(m + n), where m and n are the numbers of control points of two cross sections. It is much faster than Kehtarnavaz and De Figueiredo's merge method, whose time complexity is O(n³m²).     

Reconstruction of 3D scenes using stereoimages without rectification

The problem of reconstructing 3D scenes using stereopairs of images is considered. In contrast to traditional technologies whereby the rectification and disparity map construction are implemented, the corresponding points are determined directly on the epipolar lines that belong to the given epipolar plane. A complete three-dimensional model of an object is constructed by scanning a 3D scene using pencils of planes. The elimination of fragment-matching errors occurs by using the decimation filtering of the found points. The results of experiments that show the efficiency of the proposed technology are provided.     

Fast evolution of image manifolds and application to filtering and segmentation in 3D medical images

In many instances, numerical integration of space-scale PDEs is the most time consuming operation of image processing. This is because the scale step is limited by conditional stability of explicit schemes. We introduce the unconditionally stable semiimplicit linearized difference scheme that is fashioned after additive operator split (AOS) [Weickert, J. et al. (1998)], [Goldenberg, R et al., (2001)] for Beltrami and the subjective surface computation. The Beltrami flow [Kimmel, R. (1997) (1999)], [Sochen, N. et al. (1998)], is one of the most effective denoising algorithms in image processing. For gray-level images, we show that the flow equation can be arranged in an advection-diffusion form, revealing the edge-enhancing properties of this flow. This also suggests the application of AOS method for faster convergence. The subjective surface [Sarti, A. et al. (2002)] deals with constructing a perceptually meaningful interpretation from partial image data by mimicking the human visual system. However, initialization of the surface is critical for the final result and its main drawbacks are very slow convergence and the huge number of iterations required. We first show that the governing equation for the subjective surface flow can be rearranged in an AOS implementation, providing a near real-time solution to the shape completion problem in 2D and 3D. Then, we devise a new initialization paradigm where we first "condition" the viewpoint surface using the fast-marching algorithm. We compare the original method with our new algorithm on several examples of real 3D medical images, thus revealing the improvement achieved.     

星载SAR影像的定位与几何纠正

高精度几何纠正算法对SAR影像的广泛应用具有重要意义,而SAR影像纠正中的一个重要问题是建立影像定位模型。在阐述了距离多普勒定位模型的基础上,利用一景高分辨率的TerraSAR-X影像进行了地理定位实验,实验结果表明采用2次参数方程和3次参数方程描述卫星轨道位置所得的定位精度相差不大,且都能满足精度要求,为了计算方便和减少控制点的数目,通常可以采用二次参数方程来描述卫星位置。定位参数解算完成后,又采用了间接纠正法对影像进行了几何纠正,得出纠正精度不超过3个像素,表明该算法合理可行。     

Algorithms for matching 3D line sets

Matching two sets of lines is a basic tool that has applications in many computer vision problems such as scene registration, object recognition, motion estimation, and others. Line sets may be composed of infinitely long lines or finite length line segments. Depending on line lengths, three basic cases arise in matching sets of lines: 1) finite-finite, 2) finite-infinite, and 3) infinite-infinite. Case 2 has not been treated in the literature. For Cases 1 and 3, existing algorithms for matching 3D line sets are not completely satisfactory in that they either solve special situations, or give approximate solutions, or may not converge, or are not invariant with respect to coordinate system transforms. In this paper, we present new algorithms that solve exactly all three cases for the general situation. The algorithms are provably convergent and invariant to coordinate transforms. Experiments with synthetic and real 3D image data are reported.     

Skeleton-enhanced line drawings for 3D models

We present a novel line drawing approach for 3D models by introducing their skeleton information into the rendering process. Based on the silhouettes of the input 3D models, we first extract feature lines in geometric regions by utilizing their curvature, torsion and view-dependent information. Then, the skeletons of the models are extracted by our newly developed skeleton extraction algorithm. After that, we draw the skeleton-guided lines from non-geometric regions through the skeleton information. These lines are combined with the feature lines to render the final line drawing result using the line optimization. Experimental results show that our algorithm can render line drawings more effectively with enhanced skeletons. The resulting artistic effects can capture the local geometries as well as the global skeletons of the input 3D models.     

Lytro相机的光场图像校正与重对焦方法

目的 光场相机因其具有先拍照后对焦的能力而备受广泛关注,然而光场相机的理论研究往往受制于原始光场图像及其标定参数的获取。基于目前市面上唯一的一款消费级光场相机——Lytro相机,提出一套光场图像获取、校正和重对焦的系统化方法。方法 通过分析Lytro相机工作原理、算法和文件结构,给出了一套实现原始光场数据提取、解码、颜色校正、微透镜阵列标定和校正的完整方案。并在此基础上提出一种基于分数阶傅里叶变换的Lytro相机重对焦方法。结果 实验结果表明所提系统级方案正确,所提重对焦方法具有令人满意的重对焦性能。结论 本文Lytro相机光场图像获取和标定方法,利用分数阶傅里叶变换实现光场图像的频率域重聚焦。基于两台Lytro相机的实验结果表明,本文方法可正确获取图像数据及相机参数,重聚焦算法效果好。