多模跟踪技术在轮式侦察车图像处理器的应用

为解决目标旋转形变、遮挡、光照变化等目标跟踪的难题,对粒子滤波和尺度不变特征变换（SIFT）算法进行了改进,结合两种算法提出了决策主导模式的多模跟踪技术。该技术采用粒子滤波预测目标位置进行粗定位,SIFT特征匹配进行精定位的方法,在解决上述难题上有很好的鲁棒性。将该技术应用于轮式侦察车图像处理器,并进行了各种实验验证,结果证明了提出算法的有效性。     

Computational analysis for the determination of deleterious nsSNPs in human MTHFD1 gene

Single nucleotide polymorphisms (SNPs) are the most common genetic polymorphisms and play a major role in many inherited diseases. Methylenetetrahydrofolate dehydrogenase 1 (MTHFD1) is one of the enzymes involved in folate metabolism. In the present study, the functional and structural consequences of nsSNPs of human MTHFD1 gene was analyzed using various computational tools like SIFT, PolyPhen2, PANTHER, PROVEAN, SNAP2, nsSNPAnalyzer, PhD-SNP, SNPs&GO, I-Mutant, MuPro, ConSurf, InterPro, NCBI Conserved Domain Search tool, ModPred, SPARKS-X, RAMPAGE, FT Site and PyMol. Out of 327 nsSNPs form human MTHFD1 gene, total 45 SNPs were predicted as functionally most significant SNPs, among which 17 were highly conserved and functional, 17 were highly conserved and structural residues. Among 45 most significant SNPs, 15 were predicted to be involved in post translational modifications. The p.Gly165Arg may interfere in homodimer interface formation. The p.Asn439Lys and p.Asp445Asn may interfere in binding interactions of MTHFD1 protein with cesium cation and potassium. The two SNPs (p.Asp562Gly and p.Gly637Cys) might interfere in interactions of MTHFD1 with ligand.     

一种改进的无人机航空影像配准方法

针对目前无人机航空影像非同源、畸变大、处理量多的问题,提出一种改进的无人机航空影像配准方法。首先利用传统SIFT方法得到特征点,其次利用C均值聚类方法可实现准确的非监督分类的特点,对传统SIFT方法得到的特征点进行筛选,从而得出同名点。最后根据得到的同名点完成待匹配图像的投影变换完成配准。通过实验仿真证明该方法精度有较大提高,且可自适应处理不同图像,是一种有效的无人机航空影像匹配改良方法。     

基于几何矩预分类的无人机遥感图像自动配准方法

图像配准技术是近些年来图像处理领域发展迅速的研究方向之一.在遥感领域内,图像配准更是实现图像融合、运动检测、图像校正、图像拼接等应用的一个关键步骤.尽管国内外目前在图像配准方面提出了很多方法,但不同方法适用的图像范围不同,很多时候需要人工干预进行方法的选择.尤其对于无人机这种快速、实时获取图像的新型遥感平台,如何集合不同方法的优点以实现图像自动配准成为了关键性问题.本文在比较分析了基于SIFT和SURF特征提取图像配准方法的各自优势后,提出基于几何矩的方法对图像进行预先分类,从而决定将其分配给何种方法进行配准,实现全程自动化.实验证明,这种图像自动配准方法在拓宽了图像应用范围的同时保证了良好的配准效果.     

基于单目图像和方向的测距系统及性能分析

 针对非合作大目标的被动测距问题,介绍了一种基于单目图像序列和成像方向的测距系统,并进行了性能分析.这种测距系统的特点是无需初始距离导引,它对目标距离的估计借助于专门的定位方程,方程由目标特征线度、摄像机的空间坐标、目标在采样时刻对相机的方位角和俯仰角决定.其中,方位角和俯仰角通过光电经纬仪获得,摄像机的空间坐标通过GPS获得,目标特征线度通过相邻帧图像匹配得到.分析表明,该系统的性能主要取决于特征线度的准确度,其他测量误差如目标的方位角、俯仰角和摄像机的空间坐标对其影响较小.目标特征线度选取时,在相邻帧目标图像匹配点中,筛选出三个具有较大极限的尺度不变特征变换关键点,构造三角形及其外接圆,在外接圆上取经过三角形重心且平行或垂直于尺度不变特征变换主方向的弦线作为特征线度,这种特征线度对目标旋转不敏感.实验结果表明这种测距方法能够实现对非合作大目标的被动测距.     

基于SIFT匹配的全局运动估计方法

提出了一种基于SIFT(Scale invariant feature transform)匹配的全局运动估计算法.在SIFT初匹配的基础上,对每一个原始匹配特征点,利用所在尺度的邻域灰度信息,对其加权平均后再进行匹配,进而去除误配点.精炼后的匹配点集合作为求解全局运动参数模型的对应数据,采用最小二乘法计算模型参数.在...     

基于SIFT算法的移动机器人同时定位与地图创建

研究了基于尺度不变特征变换（SIFT）算法的移动机器人同时定位与地图创建（SLAM）方法,即在视角改变情况下,用SIFT算法对不同图像进行特征匹配,根据极线几何原理得到摄像头的旋转角度,将之与里程计的角度信息融合,从而实现较准确的自我定位与地图创建.实验表明：本方法利用电荷耦合器件（CCD）摄像头和里程计间内在的几何关系来实现SLAM,提高了低成本移动机器人的定位精度.     

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...     

Identification and structural characterization of deleterious non-synonymous single nucleotide polymorphisms in the human SKP2 gene

In SCF (Skp, Cullin, F-box) ubiquitin-protein ligase complexes, S-phase kinase 2 (SKP2) is one of the major players of F-box family, that is responsible for the degradation of several important cell regulators and tumor suppressor proteins. Despite of having significant evidence for the role of SKP2 on tumorgenesis, there is a lack of available data regarding the effect of non-synonymous polymorphisms. In this communication, the structural and functional consequences of non-synonymous single nucleotide polymorphisms (nsSNPs) of SKP2 have been reported by employing various computational approaches and molecular dynamics simulation. Initially, several computational tools like SIFT, PolyPhen-2, PredictSNP, I-Mutant 2.0 and ConSurf have been implicated in this study to explore the damaging SNPs. In total of 172 nsSNPs, 5 nsSNPs were identified as deleterious and 3 of them were predicted to be decreased the stability of protein. Guided from ConSurf analysis, P101L (rs761253702) and Y346C (rs755010517) were categorized as the highly conserved and functional disrupting mutations. Therefore, these mutations were subjected to three dimensional model building and molecular dynamics simulation study for the detailed structural consequences upon the mutations. The study revealed that P101L and Y346C mutations increased the flexibility and changed the structural dynamics. As both these mutations are located in the most functional regions of SKP2 protein, these computational insights might be helpful to consider these nsSNPs for wet-lab confirmatory analysis as well as in rationalizing future population based studies and structure based drug design against SKP2.     

An optimization method for registration and mosaicking of remote sensing images

An optimization method for registration and mosaicking of remote sensing images has been proposed. More than four pairs of control points extracted by scale invariant feature transform (SIFT) detector have been used to obtain the projective transformation relation between the reference image and text image. A method based on gradient information of the region has been proposed to assign the weighted value to the feature points. The weighted value reflects the importance of the area, which can change the result of optimization. The combination method of seam cutting and feathering technique is used to stitch two overlapped images. Our experiment shows that the alignment error of interest area is smaller than that of normal area. If the weighted parameter is given proper value, the alignment error of interest area can reduced to less than one pixel.