医学图像配准中的数据抽样方法研究

针对在基于灰度的医学图像配准中,传统数据抽样方法在过抽样时产生较多的局部极值点问题,提出了基于浮动图像灰度概率分布和其梯度信息的抽样方法.通过对人体脑部的刚体配准实验,从函数曲线和收敛性能方面,对比分析了五种数据抽样方法.实验结果表明,新抽样方法可以有效地减少局部极值点,提高归一化互信息测度的收敛性能.     

基于IPSO和综合信息的医学图像配准新方法

针对医学图像配准中采用互信息作为配准相似度函数存在配准精度不高和收敛速度慢等问题,根据图像灰度和空间结构信息,构造了一种新的基于互信息和改进型形态学梯度算子的信息配准测度函数,采用一种适用于医学图像自动配准的改进型粒子群优化(IPSO)算法,给出了一种新的基于IPSO的医学图像配准算法。实验结果表明,该配准算法稳定性好、收敛速度快,在多模态医学图像自动配准中是可行的。     

新的Jensen-Schur测度在医学图像配准中的应用

针对传统的互信息和归一化互信息运算时间长、收敛性能和抗噪声能力差的缺点,提出了一种新的医学图像配准测度,Jensen-Schur测度.从理论上分析了该测度可用于图像配准的性质,并构造了三类Jensen-Schur测度:Jensen-Schur-alpha、Jensen-Rényi-alpha、JenserrSchur-w测度.通过对人体脑部MR/CI"和MR/PET图像的刚体配准实验,从运算时间、收敛性能、抗噪鲁棒性方面,对这三类测度、互信息和归一化互信息进行了比较和分析.实验结果表明,Jensen-Schur-alpha(alpha=2或3)和Jensen-Schur-w测度的收敛性能优于其它测度,运算速度快于其它测度,对噪声有很强的鲁棒性.最后用Jensen-Sehur-alpha(alpha=2)测度进行多模态医学图像的配准实验,结果表明效果良好.     

基于改进PV插值和混合优化算法的医学图像配准

基于互信息的医学图像配准算法近来获得了广泛的应用.在计算互信息的时候,采用部分体积插值计算联合直方图无法避免在象素整数倍位移处的局部极值,有可能导致误配准.本文提出了基于B样条函数的部分体积插值方法,有效地减少象素整数倍位移处的局部极值;在优化互信息函数的时候先使用遗传算法将配准参数收敛到最优值附近,然后使用Powell方法进行局部搜索,避免误收敛.实验表明,本文提出的算法具有更好的精确性和稳定性.     

基于图像处理技术的集成电路焊接质量监测系统研究

随着计算机的发展,数字图像处理的应用为电路板焊接质量的检测提供了新的条件。重点研究了参考图像和待检测图像的图像配准算法,采用基于灰度信息的图像配准方法,使用PV插值法直接更新联合直方图并由此计算出参考图像和浮动图像的互信息值,然后采用改进的鲍威尔算法搜索最优配准参数,最后根据配准参数进行插值从而得到配准图像。最后,在缺陷检测方面通过颜色识别寻找配准图像和参考图像的差异部分,通过直方图匹配和边缘检测滤去伪缺陷和噪声部分得到缺陷部分,最后标定缺陷位置。实验结果表明,采用本文设计方法可以实现焊接缺陷的快速并且高精度的检测,从而进一步降低人工成本,提高检测效率。     

结合NGVF信息和互信息的图像配准方法

对传统的互信患配准方法进行改进,提出一种新的结合灰度信息和空间信息的配准方法.该方法分为粗略和精确配准两个阶段,首先利用图像的空间信息进行粗配准,再利用ACMI进行精确配准.文中利用的空间信息不同于一般的图像梯度信息,它是归一化的GVF矢量场信息(NGVFI),把它用于粗配准,抗噪性能良好.试验证明该方法鲁棒性很好,特别是对大噪声图像,配准效果突出,精度能达到亚像素级.     

基于简单Schur凹函数的图像配准测度研究

 使用互信息或归一化互信息进行图像配准时,由于噪声、模态、插值等影响,测度函数存在许多局部极值,收敛范围较窄,有可能导致误配准.该文根据一个简单的Schur凹函数,充分利用它的特殊上凸性来消除噪声等引起的小概率分布,并由Jensen-Schur测度、广义距离测度和 f 信息测度的定义,构造了六种新测度.从运算时间、收敛性能、抗噪鲁棒性方面,对这六种测度、互信息和归一化互信息进行了比较和分析.实验结果表明,Jensen-Schur-beta和D-beta测度的收敛性能优于其它测度,抗噪声能力强于其它测度,运算速度快于互信息和归一化互信息.     

图像插值方法对互信息局部极值的影响分析

多模态图像配准中常使用互信息作为配准度量,互信息中的联合概率密度函数一般是利用图像灰度对的统计值来代替的,而图像插值可能产生新的灰度对,造成互信息出现局部极值。该文利用一维信号从理论上分析了线性和最近邻两种插值方法对互信息的影响。理论分析表明,线性插值造成互信息局部极值的可能性较小,而最近邻插值会使互信息出现周期性局部极值。试验结果证实了该文的结论。分析结果对基于互信息的多模态图像配准具有理论参考价值。     

基于改进Hausdorff测度的医学图像配准算法

 本文针对在医学图像配准中采用Hausdorff(HD)测度算法时存在精度和计算量之间的矛盾,提出了HD测度的改进算法,在HD测度算法中首次引入非线性的Bayesian一致性函数,以达到提高配准精度和抗噪声性能的目的.实验表明,该方法有效地解决了采用传统HD距离测度算法时,因出格点和噪声点引起的误配问题,获得了较好的配准效果.     

基于JS测度和层次B样条的医学图像弹性配准

为了精确得到图像的不规则局部几何变形,提出了一种基于标记点的医学图像弹性配准新方法.首先用Jensen-Schur(JS)测度对图像进行粗配准,然后用该测度自动选取标记点对,最后用能够解决逼近精度和光滑性平衡问题的B样条细化的层次B样条插值技术来实现图像弹性配准.实验表明,本文提出的标记点自动选取方法能够产生准确的表示不规则局部变形的标记点对,所提出的配准方法能够产生光滑、准确的弹性配准变换.     

Fast and accurate image registration using Tsallis entropy and simultaneous perturbation stochastic approximation

The Tsallis measure of mutual information is combined with the simultaneous perturbation stochastic approximation algorithm to register images. It is shown that Tsallis entropy can improve registration accuracy and speed of convergence, compared with Shannon entropy, in the calculation of mutual information. Simulation results show that the new algorithm achieves up to seven times faster convergence and four times more precise registration than using a classic form of entropy.     

Further analysis of interpolation effects in mutual information-based image registration

This paper presents an analysis of the mutual information (MI) metric in rigid-body registration of two digital images, in particular, local fluctuations of the MI value due to interpolation. In contrast to existing work in this area, this paper starts with two hypothetical continuous images, based on which both sampling and interpolation effects are analyzed. This analysis indicates that an "ideal" interpolator may not be able to completely suppress the undesirable local minima of the MI metric if the sampling effect is not negligible. Several preprocessing methods are discussed for reducing the interpolation effects.     

Interpolation Artifacts in Sub-Pixel Image Registration

We consider the problem of registering (aligning) two images to sub-pixel accuracy by optimization of objective functions constructed from the images' intensity values. We show that some widely used interpolation methods can introduce multiple local optima in the energy of the interpolated image which, if not counter-balanced by other terms, can cause local optima in registration objective functions including the sum of squared differences, cross correlation, and mutual information. We discuss different solutions to address the problem based on high degree B-spline interpolation, low pass filtering the images, and stochastic integration. Numerical examples using synthetic and real signals and images are shown.     

基于区域联合Rényi熵的多模医学图像配准

 针对医学图像配准对准确性高、鲁棒性强和速度快的要求,本文提出一种新的基于区域联合Rényi熵的多模配准算法.该算法将区域信息融入到联合Rényi熵中,并使用最小生成树来估计区域联合Rényi熵.这样,不仅改善了传统配准方法由于忽略像素空间信息造成的配准鲁棒性的降低,而且避免了使用直方图估计高维熵遇到的"维数灾难"问题.实验结果表明在图像含有噪声、灰度不均匀和初始误配范围较大的情况下,该算法在达到良好配准精度的同时,具有鲁棒性强、速度快的优点.作为一种一般性的配准算法,基于区域联合Rényi熵的配准方法还可以应用到图像配准以外的更广阔的领域,如图像检索、对象识别等.     

An automatic nonrigid registration for stained histological sections.

Automatic computer-based analyses of histological sections which are differently stained require that they are related to each other. Most registration methods are only able to perform rigid-body motion and are sensitive to noise and artifacts. Histological images, however, are accompanied by several artifacts and different contrasts, which require a nonrigid registration. In this paper, we present a hierarchical nonrigid registration algorithm able to align images, which contain minor image artifacts. The algorithm requires no a priori knowledge of the true image. The hierarchical design of the algorithm enhances robustness and accuracy, and saves computational costs. The proposed algorithm is decomposed into a fast, coarse, rigid registration step and a slower, but finer, nonrigid step. For the coarse registration, we use image pyramids, while for the second step, we combine a point-based registration with an elastic thin-plate spline interpolation. Accuracy tests, performed for 20 histological images obtained from human arteries, have shown that the error measure is acceptable, and that the image noise does not cause a problem. The associated convergence rate of the mean pixel displacement error during the rigid and nonrigid registrations is satisfying. The algorithm can be applied to various multicontrast elastic registration problems in medical imaging and may be extended to three dimensions.     

A high-resolution interpolator for incremental encoders based onthe quadrature PLL method

This paper presents a new high-resolution interpolator for incremental encoders based on the quadrature phase-locked loop method proposed by Emura. Until now, this method has been applied to controllers of high-precision servomechanisms and has shown excellent performance. In this paper, the authors apply the method for the first time to an interpolator for incremental encoders. The experimental results show that high-speed interpolation is possible, with a maximum output frequency of 25 MHz-20 times higher than that of conventional interpolators. The interpolator proposed was also tested for noise rejection with a high-speed numerically controlled gear grinding machine, and has shown good noise rejection capability. This paper presents the design of the interpolator and the experimental results     

Interpolation of complex stationary processes

The problem of minimum mean-square infinite extent interpolation for discrete-time stationary complex stochastic processes is studied. The interpolator consists of linear combinations of samples of the process and of their complex conjugate. The expressions of the interpolator and of the approximation error are derived and various consequences are examined. It is shown in particular that the approximation error may be zero while the interpolation error obtained when using only linear combinations of the samples is maximum