集成SIFT算法与检测模型优化的UAV影像匹配方法

低空无人机(UAV)测量凭借着低成本、高效率、高精度的数据采集模式,可快速获取高空间分辨率的影像数据,已经成为遥感领域的一种重要技术手段.其中,影像匹配技术是UAV影像数据处理的重要步骤,图像间的匹配直接影响后期三维场景的精度及视觉效果.针对高原山地的高差起伏变化大地形复杂,植被覆被率高及地物分布不规则等问题存在,致使...     

基于CSIFT的彩色图像配准技术研究

图像配准在计算机视觉、遥感、医学诊断与治疗、环境监测等领域有广泛的研究应用.目前,多数算法是将彩色图像转化为灰度图后再配准,色彩信息的丢失可能会引起误配准.为此,提出一种基于CSIFT(Colored scale invariant feature transform)的彩色图像配准方法,求出彩色图像各个位置处的颜色不变最,以颜色不变量作为输入图像,再提取特征点并描述特征点周围的信息,通过最近邻匹配法求出图像问的匹配对,最后利用匹配的特征求取图像间的变换参数及配准后图像.实验结果表明,对彩色图像进行已知参数值变换时,该算法能得到精度高、误差小的计算结果;对变换关系末知的彩色图像,也能准确地求出图像间的映射关系;且多数情况下运行速度较SIFT(Scale invariant feature transform)快.     

基于特征点及优化理论的图像自动拼接方法

提出了一种新的图像拼接方法,首先利用相位一致性(phase congruency)算法进行特征点检测,利用本文提出的匹配点优选策略进行特征点对自动选取,然后用LM(Levenberg-Marquardt)算法进一步优化变换矩阵,最后对拼接结果进行融合处理,获得无缝拼接的图像.该方法把基于特征点和基于优化理论的拼接方法有效相结合,且能充分利用图像重叠部分的信息,在一定程度上克服了噪声及光照不均的影响,较传统方法具有更强的鲁棒性和更高的拼接精确度.试验结果证明了该方法的有效性.     

基于改进SIFT特征点匹配的图像拼接算法

基于传统SIFT方法和图像像素加权平滑融合的思想,提出了一种改进SIFT特征点的图像拼接方法。该方法首先利用Canny边缘检测算法获得图像的边缘点坐标,通过和SIFT算法关键点坐标进行对比,去除不稳定响应点;其次通过K-L变换降低算法复杂度,对得到的匹配点对,使用RANSAC算法进行提纯,计算投影变换模型参数;最后使用渐入渐出的加权融合算法平滑图像,消除图像之间的拼接缝隙。该算法的可行性和有效性通过实验结果可以得到证明。     

基于非抽样剪切波的加权区域图像融合

以提升遥感图像和多聚焦图像的融合精度为目的,结合非下采样剪切波变换（NSST）可以捕捉图像的细节特征,提出了一种NSST和加权区域特性的图像融合方法。利用非下采样剪切波变换对源图像进行多尺度、多方向分解,得到低频子带和高频子带,低频子带系数采用改进梯度投影的非负矩阵分解（NMF）,高频子带系数采用加权区域能量和区域方差相结合的融合策略,然后应用非下采样剪切波的逆变换得到融合的图像。实验结果表明:该方法从主观视觉方面很好地保留了多幅图像的有用信息,给出该方法与其他融合算法在客观评价指标应用信息熵EN、互信息MI和加权边缘信息量QAB/F的比较结果 。     

改进SIFT变换与客观评价结合的图像配准算法

针对SIFT(Scale Invariant Feature Transform)算子在大幅复杂图像中提取的过多不稳定特征点及在只有少量重合区域下图像配准过程中出现的过多误匹配,导致图像配准精度下降。提出一种改进的SIFT算法,在对目标图像提取SIFT特征后,利用双向BBF(Best-Bin-First)匹配算法对提取的特征点进行匹配,采用SIFT描述子的尺度以及梯度方向信息建立最小邻域匹配剔除误匹配点,通过随机抽取一致性算法(RANSAC)进一步筛选匹配点,并利用最小二乘法结合多项式近似拟合出变换模型,利用局部均方根有效值(RMS)评价映射矩阵与实际图像的误差,找出并删除引起误差的误匹配点,迭代至配准图像符合评价标准后,计算出精确变换模型。实验结果表明,该算法提高了大幅复杂图像在少量重合区域时的配准精度。     

图像局部特征自适应的快速SIFT图像拼接方法

针对目前图像拼接中计算量较大、实时性较差的问题,本文提出了一种图像局部特征自适应的快速尺度不变特征变换(SIFT)拼接方法。首先,对待拼接图像分块,确定图像局部块的特征类型;接着自适应采用不同的简化方法提取各局部块的特征点。然后,通过特征匹配求出变换矩阵,并结合RANSAC算法去除伪匹配对。最后,通过图像融合得到最终的拼接图像。文中使用提出的方法对3组待拼接图像进行实验。从实验结果可以看出:与标准拼接方法相比,本文改进方法的计算速度提升了30%~45%。因此,这种方法能够在保证图像拼接质量的前提下,有效提高图像拼接的效率,克服图像拼接中计算复杂度高的问题,在实际图像拼接中具有一定的应用价值。     

基于空间广义相干系数加权的医学合成孔径超声成像

为解决广义相干系数用于合成孔径成像中所存在的运算量大,图像对比度改善有限等问题,提出空间广义相干系数加权成像方法。该算法根据单个孔径成像结果之间的相干性来计算相干系数,通过加权空间合成进行成像。采用对FieldⅡ仿真点目标和吸声斑目标的数据进行成像表明,算法不仅使运算量减少N(N为阵元数)倍,而且相对于传统的广义相干系数算法,算法对散射点成像可提高信噪比7 dB,对于吸声斑成像可提高对比度3.2 dB。可见算法可以有效地提高成像速度和成像质量。      

基于量子粒子群优化的正交小波加权多模盲均衡算法

针对常数模盲均衡算法(CMA)均衡高阶正交振幅调制信号(QAM)存在收敛速度慢、稳态误差大的缺点, 提出了基于量子粒子群优化的正交小波加权多模盲均衡算法(QPSO-WTWMMA). 该算法根据高阶QAM信号星座图分布特点, 将量子粒子群优化算法(QPSO) 和正交小波变换融入于加权多模盲均衡算法(WMMA)中. 因而, 利用QPSO对均衡器权向量进行了优化, 利用正交小波变换降低了输入信号的自相关性, 利用WMMA选择了合适的误差模型匹配QAM星座图. 理论分析及水声信道仿真结果表明, QPSO-WTWMMA算法可以获得更快的收敛速度和更低的稳态误差, 在水声通信中具有重要的参考价值.     

基于图像融合的动态轮廓线跟踪新方法

红外与可见光传感器是目标跟踪识别系统中常用的两种传感器,对这两种传感器图像进行融合能有效提高系统跟踪检测的准确性。将动态轮廓线模型与图像融合结合,在特征搜索过程中利用特征点准确地完成了图像配准,同时使用了一种新的特征级融合方法,将两种图像中目标轮廓的B样条曲线控制点进行实时微分耦合。这种耦合将Curwen提出的微分耦合机制作了改进,利用图像配准把刚性硬模板改变为实时的变换模板并推导了融合后动态轮廓线的新的动力学方程。这种融合利用了红外图像目标轮廓信息约束可见光图像中动态轮廓线的收敛形状,有效地提高了可见光图像目标跟踪的准确性。对运动人手序列图像的对比跟踪实验表明,这种融合使得可见光图像中动态轮廓线平均跟踪误差减小了60．25％。     

Combining Sparse and Dense Features to Improve Multi-Modal Registration for Brain DTI Images

A new solution to overcome the constraints of multimodality medical intra-subject image registration is proposed, using the mutual information (MI) of image histogram-oriented gradients as a new matching criterion. We present a rigid, multi-modal image registration algorithm based on linear transformation and oriented gradients for the alignment of T2-weighted (T2w) images (as a fixed reference) and diffusion tensor imaging (DTI) (b-values of 500 and 1250 s/mm²) as floating images of three patients to compensate for the motion during the acquisition process. Diffusion MRI is very sensitive to motion, especially when the intensity and duration of the gradient pulses (characterized by the b-value) increases. The proposed method relies on the whole brain surface and addresses the variability of anatomical features into an image stack. The sparse features refer to corners detected using the Harris corner detector operator, while dense features use all image pixels through the image histogram of oriented gradients (HOG) as a measure of the degree of statistical dependence between a pair of registered images. HOG as a dense feature is focused on the structure and extracts the oriented gradient image in the x and y directions. MI is used as an objective function for the optimization process. The entropy functions and joint entropy function are determined using the HOGs data. To determine the best image transformation, the fiducial registration error (FRE) measure is used. We compare the results against the MI-based intensities results computed using a statistical intensity relationship between corresponding pixels in source and target images. Our approach, which is devoted to the whole brain, shows improved registration accuracy, robustness, and computational cost compared with the registration algorithms, which use anatomical features or regions of interest areas with specific neuroanatomy. Despite the supplementary HOG computation task, the computation time is comparable for MI-based intensities and MI-based HOG methods.     

On the robustness of Harris detector in image watermarking attacks

This paper investigates the misplacement issue presented on the detection of feature points by using the well-known Harris detector, under several watermarking attacks. Harris detector, as a powerful tool in finding image's points of high importance, called “interest points”, has been widely used to locate specific image portions where watermarks are embedded, according to some watermarking insertion procedure. Although Harris detector constitutes a significant part of a typical feature-based watermarking process, no work dealing with its performance under certain attacks has been reported yet. This work studies the accuracy of Harris detector in finding the feature points, which define the watermarked regions of attacked images, firstly as an individual processing module and secondly as a part of a feature-based watermarking methodology. Detailed theoretical analysis followed by appropriate simulation experiments has shown that the incorporation of the Harris detector into the watermarking procedure presents misplacement issues that can lead the watermarking algorithm to unreliable results. In the way of decreasing these misplacement errors, a novel algorithm that extracts more stable interest points, with promising performance is also proposed.     

Sphere-based calibration method for trinocular vision sensor

A new method to calibrate a trinocular vision sensor is proposed and two main tasks are finished in this paper, i.e. to determine the transformation matrix between each two cameras and the trifocal tensor of the trinocular vision sensor. A flexible sphere target with several spherical circles is designed. As the isotropy of a sphere, trifocal tensor of the three cameras can be determined exactly from the feature on the sphere target. Then the fundamental matrix between each two cameras can be obtained. Easily, compatible rotation matrix and translation matrix can be deduced base on the singular value decomposition of the fundamental matrix. In our proposed calibration method, image points are not requested one-to-one correspondence. When image points locates in the same feature are obtained, the transformation matrix between each two cameras with the trifocal tensor of trinocular vision sensor can be determined. Experiment results show that the proposed calibration method can obtain precise results, including measurement and matching results. The root mean square error of distance is 0.026 mm with regard to the view field of about 200×200 mm and the feature matching of three images is strict. As a sphere projection is not concerned with its orientation, the calibration method is robust and with an easy operation. Moreover, our calibration method also provides a new approach to obtain the trifocal tensor.     

一种基于特征点间线段倾角的姿态测量方法

基于目标特征点间线段倾角信息,提出了一种适合于目标远距离成像和相机内参未知条件下解算目标姿态的目标3维姿态测量方法.采用仿真图像对该方法的正确性进行了验证.实验结果:姿态测量误差绝对值均值小于0.6°,且目标成像尺寸为350pixel时,姿态测量误差绝对值小于0.5°.实验表明该算法具有较高解算准确度和较强的收敛性.     

基于优势特征图像融合的水下光学图像增强

针对水下光学图像颜色失真、非均匀光照、对比度低的问题,提出基于优势特征图像融合的水下光学图像增强算法.首先,提出改进的暗通道先验算法去除退化图像中的不均匀浑浊并均衡色彩;其次,对颜色校正图像分别使用基于加权分布的自适应伽玛校正算法和限制对比度自适应直方图均衡-同态滤波算法,增强颜色校正图像对比度并使其亮度均衡;最后,定义三幅融合图像即颜色校正图像、亮度均衡图像、对比度增强图像的关联权重图,通过多尺度融合算法获得融合图像.与单一预处理算法只能解决对应的退化现象相比,该算法对单幅退化图像进行多算法处理,得到三幅优势特征图像,通过不同权重的组合最大程度地将各优势特征相结合,得到的综合效果远超各单一算法优化效果,不再局限于解决颜色失真等单一问题.将本文算法与现有算法在主观评价和客观评价两方面进行实验对比,结果表明,该算法可以有效平衡水下图像的色度、饱和度及清晰度,视觉效果接近自然场景下的图像.     

基于改进空间频率域采样的天文光干涉望远镜阵列优化

提出一种基于改进空间频率域(UV)采样的阵列评价函数,用于长基线天文光干涉望远镜阵列几何结构的优化。该评价函数将UV采样区域沿径向和角度方向分别进行划分,统计划分所得区域中UV采样点数目并计算UV采样点密度,以UV采样点密度偏离理想高斯分布的大小作为评价依据。在具体的优化技术上,利用遗传算法的全局收敛特性,降低了传统算法对初始结构的依赖,采用该评价函数对6孔径望远镜阵列进行优化设计,并与国际主流天文光干涉阵列CHARA进行了性能对比。分析结果表明:优化所得Array-6阵列UV采样点密度分布具有径向连续覆盖和低频强调的特点,有利于对轮廓信息的恢复;双星模拟成像实验中Array-6阵列重构图像相对于原始图像的误差为21.34,相比CHARA阵列降低了18.16%,具有更高的成像质量。该优化算法具备优化大孔径数目阵列的能力,对于射电波段望远镜阵列的优化设计亦有一定的参考意义。     

Kernel-aligned multi-view canonical correlation analysis for image recognition

Existing kernel-based correlation analysis methods mainly adopt a single kernel in each view. However, only a single kernel is usually insufficient to characterize nonlinear distribution information of a view. To solve the problem, we transform each original feature vector into a 2-dimensional feature matrix by means of kernel alignment, and then propose a novel kernel-aligned multi-view canonical correlation analysis (KAMCCA) method on the basis of the feature matrices. Our proposed method can simultaneously employ multiple kernels to better capture the nonlinear distribution information of each view, so that correlation features learned by KAMCCA can have well discriminating power in real-world image recognition. Extensive experiments are designed on five real-world image datasets, including NIR face images, thermal face images, visible face images, handwritten digit images, and object images. Promising experimental results on the datasets have manifested the effectiveness of our proposed method.     

尺度不变特征与几何特征融合的人耳识别方法

要提高人耳的识别率,关键是特征的提取与表达.尺度不变特征变换(SIFT)技术是局部点特征提取算法,在尺度空间寻找极值点,提取对图像的尺度和旋转变化具有不变性,对光照变化和图像变形具有较强的适应性的特征向量.尝试用SIFT技术来提取外耳图像的结构特征点以形成稳定的特征描述子,为了克服一幅图像中有多个局部描述子相似的问题,在SIFT特征描述子中融入一个耳廓几何特征.最后采用特征向量的欧氏距离作为两幅图像相似性度量标准进行人耳识别.在耳图像库七进行实验.结果表明,该方法不仅可以有效地提取人耳特征,通过少量特征可获得较高的识别率,而且对耳图像刚体变化具有较强的稳健性.     

基于曲率滤波优化的遥感影像去雾算法

针对山区容易积聚云雾不利于遥感影像解译,且遥感影像因包含大量信息数据计算速率慢,在去雾过程中遇到雾量过多的情况去除效果差的问题,提出了一种曲率滤波优化与非局部去雾方法相结合的光学遥感影像去雾算法.首先通过曲率滤波降低了图像的总能量,从而减少了图像占用的内存,提高了计算效率.然后通过霍夫变换表决机制筛选出符合的大气光值,...