网格形变细分的大视差图像拼接算法

针对大视差图像拼接后重叠区域出现重影、非重叠区域发生透视失真等问题,提出一种改进的大视差图像拼接算法。利用尽可能投影算法(APAP)建立低密度网格形变,根据待拼接图像成对匹配点的分布对重叠区域内的网格形变进行细分。通过随机采样一致性算法计算全局最优相似矩阵,校正非重叠区域发生的透视失真现象。将全局最优相似矩阵与网格单应矩阵加权叠加,实现目标图像形变。在此基础上,对目标图像重叠区域进行内容感知,保留重要度较低的区域并完成拼接,以避免重叠区域出现重影问题。实验结果表明,相对APAP、SPHP等算法,该算法的拼接效果更能还原真实场景,且拼接图像的均方根误差值较低。     

混合样条大尺度医学图像弹性配准算法研究*

针对大尺度形变医学图像配准速度慢和精度低的特点,提出一种结合薄板样条(TPS)和B样条的弹性配准方法。该方法采用尺度不变特征变换算法(SIFT)进行图像特征提取与匹配,利用TPS算法将特征点对作为输入进行预处理,以降低浮动图像的形变尺度,从而提高下一步B样条配准的速度与精度。然后使用局部区域细化层次B样条方法将TPS生成的较稀疏的形变网格作为初始网格,结合有限记忆优化算法(L-BFGS)对控制网格做进一步地处理,此过程只对形变较大的局部区域进行细化,以实现与参考图像的快速精确配准。实验结果表明,该方法较层次B样条方法有效地提高了配准的速度和精度。     

基于改进网格划分统计的特征点快速匹配方法

针对图像特征点匹配算法大多存在数据量大和计算耗时长等问题,提出一种改进网格划分统计的特征点快速匹配算法。首先将图像的长宽比作为约束项,把图像划分成多个非重叠的方形状网格,并统计网格内的粗匹配特征点数量,然后利用改进的五宫格统计方法剔除错误匹配,即将特征点所在网格的相邻对称的四个网格作为邻域范围,把五宫格特征分数与新提出的阈值公式计算的值进行比较,最终得到精匹配特征点集;在OxFord数据集和实际拍摄的无人机遥感图像上,将本文算法与多种算法进行比较,实验结果表明,该方法在保证精确率和召回率接近当前最新的特征点快速匹配算法的情况下,运行速度相对提高了35.6 %,证明了特征点匹配的实时性和有效性。     

基于改进SFM的三维重建算法研究

针对现有运动恢复结构算法重建模型存在点云稀疏等问题,提出一种利用不同匹配数据进行模型重建的算法。首先通过对比上下文直方图(CCH)生成匹配数据,利用M估计抽样一致(MSAC)估算图像基础矩阵,进而分解得到平移和旋转矩阵,并根据相机内参计算投影矩阵,然后利用KLT匹配算法更新匹配数据,最后三角化生成三维点云。该算法匹配精度高,图像基础矩阵易于收敛,通过位移实现特征点匹配,弥补了图像低频区域匹配数据不足的缺陷。实验结果表明,与现有算法相比,该算法生成的点云更致密;在真实环境下,该算法可用于物体三维重建。     

基于特征点加细的多分辨率人脸形变模型及人脸建模

提出基于特征点加细的原型三维人脸对应方法建立多分辨人脸形变模型,并根据该形变模型的特点使用多分辨模型匹配方法由单张正面人脸图像进行三维人脸建模。该方法以人脸模型上的眼、眉、口、鼻等主要几何特征为基准点标注基础网格,然后通过加细特征点网格完成原型人脸之间的对应,进而建立多分辨率的形变模型;根据形变模型的构造特点,把待匹配图像按照与模型相同方式进行加细,然后进行多分辨的人脸模型匹配。实验结果表明,新的对应算法可以有效地实现原型三维人脸之间的对应,能够克服传统的光流对应算法对应效果差,算法精度低的缺陷,提高形变模型的精度。新的匹配算法不仅能够加速模型的匹配速度,而且可提高模型匹配的效率和精度,缩短模型匹配的时间。     

静止卫星图像配准中海陆边界模板的生成方法

在使用地标进行遥感图像导航与配准时,以海陆边界为特征进行匹配操作是一种常用的方法,尤其适合昼夜图像灰度值变化剧烈的红外通道遥感图像。将海陆边界模板图像作为匹配标准,其精确性对于导航与配准的精度具有决定性的作用。针对静止卫星图像配准过程的特点,对海陆边界模板的生成方法进行研究,提出一种静止卫星图像配准中海陆边界模板的生成方法,将生成过程分为目标网格生成、全球海岸线数据库选取及查找算法3个部分,通过不断挖掘数据特点给出3种查找算法,并对各算法的精度和效率进行分析。实验结果表明,该方法生成的海陆边界模板与主流工具的生成结果有90%以上的匹配度,且算法效率较高,具有良好的应用价值。     

一种基于两步运动估计的系列图像匹配算法*

提出了一种基于两步运动估计的系列图像匹配算法。首先运用简化仿射运动估计对相邻两幅图像进行粗略匹配,然后运用改进的光流运动估计得到比较精确的匹配图像;此外,在保证简化仿射变换参数正确的同时,使用了更加简单直接的解法,大大降低了计算的复杂性。该算法弥补了传统图像匹配方法的不足,而且使图像匹配残差明显减小。实验表明,对存在大位移运动、复杂形变的图像,该算法更有效。     

基于自适应网格矢量编码的人脸快速匹配算法

在分析传统图像匹配算法优缺点的基础上,提出了一种基于自适应网格矢量编码的人脸快速匹配算法.该算法首先对人脸进行二值化边缘分割和脸部特征轮廓提取,并对图像的细节边缘和特征轮廓进行自适应网格划分,然后对网格内的边缘段和特征轮廓段进行链码描述及矢量编码,最后通过计算编码后的方向矢量矩阵相似度来实现人脸图像的匹配识别.通过对ORL标准人脸图像库的实验仿真及对比结果表明,该算法在降低计算复杂度及提高识别精度的同时,有效地解决了由图像旋转、姿态变化引起的微小形变以及不同光照条件所带来的图像识别率低的问题.     

基于动态规划的城市道路提取改进算法

从航空或卫星影像中提取道路一直是研究的热点,基于动态规划的道路提取算法是最有效的算法之一。该文基于LiDAR点云数据特征改进了该算法的代价函数,进而提高了基于动态规划的道路提取算法的鲁棒性。为正确地融合航拍图像和LiDAR点云数据,研究了航拍图像和LiDAR点云数据的匹配算法。最后,通过试验验证了算法的正确性。     

基于形变模型的三维人脸重建方法及其改进

形变模型（morphable model）是近几年出现的三维人脸建模新方法．该方法使用原型人脸的组合表示新的人脸,对于特定人脸图像,通过模型匹配实现了三维人脸的自动重建．虽然形变模型具有自动化、真实感好等优点,但现有形变模型的建立依赖于不稳定的人脸图像对应光流算法,模型匹配只考虑了一般光照环境下的人脸重建问题,且建模计算量大．针对以上问题,文章对形变模型进行了改进：提出了网格重采样的方法,实现了模型人脸数据的精确对应;建立了多分辨率的三维人脸模型;在模型匹配过程中采用了多光源光照模型,使模型可适用于复杂光照环境下的人脸重建．实验结果表明,上述改进可以有效提高模型匹配的效率和准确性以及模型对光照的适应性．     

A physically based satellite rainfall estimation method using fluid dynamics modelling

A cloud motion winds (CMW) method is presented for improving quantitative rainfall estimation advection schemes that use both infrared (IR) and passive microwave (PMW) satellite data. Advection schemes are used to provide quantitative rainfall estimates by combining more direct PMW rainfall estimates with more frequent IR cloud top temperature measures using a two‐step technique: (1) PMW estimates are transported along CMW trajectories calculated with an advection scheme at subpixel resolution; and (2) PMW estimates are calibrated using the IR gradient along those trajectories. These schemes outperform traditional methods of satellite rainfall estimation but no clear physical basis for the procedure has yet been described. Here, the physical basis for the image processing techniques used in advection techniques is described. It is shown that geostationary satellite‐derived CMW from IR sensors can be modelled in terms of fluid dynamics using Navier–Stokes equations. This approach allows for modelling the problem as equivalent to the flow of a brightness temperature field, also providing subpixel resolution and unlimited rotation/deformation possibilities. The method is illustrated with rainfall estimates from a numerical weather prediction (NWP) model and with 3‐hourly PMW products as simulation data, obtaining consistent results.     

高速水面艇视觉系统电子稳像算法

针对高速水面艇视觉系统在采集视频过程中, 由于高速运行、水流运动和风力影响等因素造成的视频图像抖动问题, 根据高速水面艇视频图像运动特点, 例如同时含有平移、旋转和变焦运动等, 采用尺度不变特征变换算法提取视频图像中的特征点, 利用仿射模型求解运动参数, 运用Kalman滤波对视频图像中的正常扫描进行滤波, 最后用相邻帧补偿法对每帧图像进行补偿, 实现高速水面艇的视频图像稳像处理。算法用于高速水面遥控艇采集到的视频上进行对比验证分析, 结果表明算法对高速水面艇视觉系统下的视频图像稳像处理快速、有效。     

基于二维透射变换的视频全景图拼接算法

视频序列全景图拼接技术是视频图像分析的关键技术之一。针对帧间运动模型为二维投影变换的常见视频序列,提出了一种全景图拼接算法.该算法首先采用改进的基于RANSAC的特征点匹配算法对相邻帧的运动参数进行估计,接着采用直接法对各帧与全景图间的运动参数进行精确调整,最后运用自适应中值滤波绘制出全景图。实验结果表明,该方法能够生成较高质量的全景图,基本能实现无缝连接。     

Cloud detection from thermal infrared images using a segmentation technique

In this paper, a fast and accurate method is proposed for two cloud detection tests on thermal infrared (IR) data obtained from satellite images over sea; an IR gross cloud test and a spatial coherence test. The proposed method is based on a regional segmentation technique. After the segmentation of an IR image, small regions were regarded as cloudy due to their high spatial variability in temperature. This technique preserved the spatial resolution of the detected cloud image which would be degraded by the conventional spatial coherence test. It also reduced the computation dramatically, compared to the conventional spatial coherence test. An accurate temperature threshold between clear sea and clouds was determined directly from the segmented image. This post-determined threshold was found to be more accurate than pre-determined temperature thresholds. Since this algorithm does not require any human interaction, it can be combined with other tests in an automatic cloud detection algorithm.     

Pipelining parallel image compositing and delivery for efficient remote visualization

Scientific datasets of large volumes generated by next-generation computational sciences need to be transferred and processed for remote visualization and distributed collaboration among a geographically dispersed team of scientists. Parallel visualization using high-performance computing facilities is a typical approach to processing such increasingly large datasets. We propose an optimized image compositing scheme with linear pipeline and adaptive transport to support efficient image delivery to a remote client. The proposed scheme arranges an arbitrary number of parallel processors within a cluster in a linear order and divides the image into a carefully selected number of segments, which flow through the linear in-cluster pipeline and wide-area networks to the remote client consecutively. We analytically determine the segment size that minimizes the final image display time and derive the conditions where the proposed image compositing and delivery scheme outperforms the traditional schemes including the binary swap algorithm. In order to match the transport throughput for image delivery over wide-area networks to the pipelining rate for image compositing within the cluster, we design a class of transport protocols using stochastic approximation methods that are able to stabilize the data flow at a target rate. The experimental results from remote visualization of large-scale scientific datasets justify the correctness of our theoretical analysis and illustrate the superior performances of the proposed method.     

A novel nonlinear algorithm for typhoon cloud image enhancement

A novel nonlinear gray transform method is proposed to enhance the contrast of a typhoon cloud image. Generally, the typhoon cloud image obtained by a satellite cannot be directly used to make an accurate prediction of the typhoon’s center or intensity because the contrast of the received typhoon cloud image may be bad. Our aim is to extrude the typhoon’s eye in the typhoon cloud image. A normalized arc-tangent transformation operation is designed to enhance global contrast of the typhoon cloud image. Differential evolution algorithm is used to choose the optimal nonlinear transform parameter. Finally, geodesic activity contour model is used to extract the typhoon’s eye to verify the performance of the proposed method. Experimental results show that the proposed method can efficiently enhance the global contrast of the typhoon cloud image while greatly extruding the typhoon’s eye.     

Multi-channel satellite cloud image fusion in the tetrolet transform domain

A novel multi-channel satellite cloud image fusion algorithm constructed in the tetrolet transform domain is proposed. Tetrolet is successfully applied in image denoising, image sparse representation, and image restoration. In this paper, tetrolet transform was introduced into the field of satellite cloud image fusion since its sparse degree is high. Tetrolet can describe the geometric structure feature of the satellite cloud image very well. First, tetrolet transform must be implemented into the multi-channel satellite cloud images to obtain low- and high-frequency coefficients and corresponding covering distribution values. Then, a Laplacian pyramid algorithm must be used to decompose the low-frequency portion in the tetrolet domain by averaging the values of its top layer and taking the maximum absolute values of the other layers. While reconstruction is implemented in this stage, the algorithm takes the maximum standard deviation of the high-frequency parts for each block in the tetrolet domain. Last, an inverse tetrolet transform must be used to obtain the final fused image. This paper compares the proposed image fusion algorithm to three similar image fusion algorithms: the curvelet image fusion algorithm, the non-subsampled contourlet transform (NSCT) image fusion algorithm, and the tetrolet image fusion algorithm. Mutual information, joint entropy, mean structural similarity (MSSIM), standard deviation, and average relative deviation are used as objective criteria to evaluate the quality of the fused images. In order to verify the efficiency of the proposed algorithm, the fusion cloud image is used to determine the centre location of eye and non-eye typhoons. Experimental results show that the proposed algorithm performs well when fusing the information in multi-channel satellite cloud images and improves the precision of locating the typhoon’s centre. The proposed algorithm’s comprehensive performance is superior to similar image fusion algorithms.     

飞行器表面矢量场数据动态纹理可视化方法

面向飞行器表面流场数据可视化的应用需求,提出一种基于线性卷积（LIC）及纹理平流（IBFVS）相结合的动态纹理可视化方法。算法通过将IBFVS方法的背景随机噪声替换为LIC纹理方式,结合了LIC纹理结果对比度高及IBFVS方法生成速度快的优势;针对LIC绘制速度慢的不足,利用GPU对曲面矢量场投影并插值,生成规则矢量数据场;用GPU对LIC部分进行并行加速,有效提高了LIC纹理图像产生速度;将LIC结果图像加入到IBFVS进行平流,生成纹理图像,最后加入颜色映射,丰富流场信息。实验结果表明,该方法生成的飞行器表面动态纹理图像对比度高,清晰度强,实时绘制性能好。     

Optical flow approximation based motion object extraction for MPEG-2 video stream

This paper presents a compressed-domain motion object extraction algorithm based on optical flow approximation for MPEG-2 video stream. The discrete cosine transform (DCT) coefficients of P and B frames are estimated to reconstruct DC + 2AC image using their motion vectors and the DCT coefficients in I frames, which can be directly extracted from MPEG-2 compressed domain. Initial optical flow is estimated with Black’s optical flow estimation framework, in which DC image is substituted by DC + 2AC image to provide more intensity information. A high confidence measure is exploited to generate dense and accurate motion vector field by removing noisy and false motion vectors. Global motion estimation and iterative rejection are further utilized to separate foreground and background motion vectors. Region growing with automatic seed selection is performed to extract accurate object boundary by motion consistency model. The object boundary is further refined by partially decoding the boundary blocks to improve the accuracy. Experimental results on several test sequences demonstrate that the proposed approach can achieve compressed-domain video object extraction for MPEG-2 video stream in CIF format with real-time performance.     

遥感视频实时云检测的FPGA高效实现方法

针对目前遥感图像云检测算法及算法运行所需硬件平台复杂度高,无法进行在轨实时检测的问题,提出了一种基于FPGA的面向卫星在轨实时运行的遥感视频云检测方法.首先根据不同的遥感视频输入格式对其自适应降采样处理;其次对顺序流入的图像自适应阈值分割,然后对分割后的图像进行聚类获取云区域,进而提取每一块云区域的特征向量;最后计算整幅图像的云覆盖率和可用度,以此判断是否将图像下传.实验结果表明,在60 MHz的时钟下,且Camera Link接口每个时钟周期同时输入两个像素时,822×1096大小的遥感视频云检测速度可达132 fps,相对于传统的嵌入式双核CPU,速度提升了6～7倍.该方法可实现卫星在轨实时云检测,极大地缓解了有限的星地数传带宽和巨大的遥感数据量之间的矛盾,大幅提升遥感卫星系统应用效能,具有很强的实用价值.