铁路运行环境下的车载相机姿态估计

基于对铁路运行环境中所拍摄的图像特征的分析,提出了一种修正的线段检测子(MLSD),并利用该线段检测子结合最小二乘拟合及交叉迭代优化方法对消失点进行快速检测,进而利用消失点坐标实现了相机姿态的估计,为后续调整机器视觉算法进行铁路运行环境的场景重建提供重要依据。运用该方法,基于车载相机所拍摄的图像对车载相机姿态进行估计,实验所得的估计值满足了车载检测系统的精度要求,表明了此方法对于估计相机姿态有效的。     

一种新的红外成像末制导目标跟踪方法

为了稳定跟踪导弹末制导阶段的红外目标,提出了一种基于尺度不变特征变换的红外目标跟踪算法.尺度不变性特征变换所提取的图像纹理特征具有尺度和旋转不变性,跟踪算法分别提取目标模板和待跟踪图像的尺度不变特征变换特征.根据最小欧氏距离准则提取目标模板与待跟踪图像间相匹配的尺度不变特征变换特征点对,利用该特征点对拟合反映两图像间映射关系的仿射模型,并据此估计目标中心位置及调整目标模板尺寸.仿真结果表明,跟踪算法能够较好地实现在导弹末制导阶段对红外地面杂波背景下目标的稳定跟踪,其跟踪准确度和稳定度优于传统方法.
关键词:末制导跟踪|尺度不变性特征变换|特征匹配|仿射模型     

综合诊断光路快速自动准直系统设计与实现

为了实现综合诊断光路全程快速准直,从并行处理、快速收敛电机运动模型和最小外切圆特征点提取三个方面对自动准直系统进行了总体设计.首先,对全部准直步骤采取并行、串并结合的方式执行,并将关键准直任务归纳成单元准直模型;其次,使用准直数学模型,实现单元准直模型中电机的快速收敛;最后,针对小孔图像灰度对比低、分布极不均匀、光斑很不完整的特点,提出在最小外切圆的圆环上,取两两距离最大的100个点为特征点,使用最小二乘法进行圆拟合法计算小孔图像圆心.研究表明优化后系统的准直时间从优化前40min减少到8min,小孔图像圆心误差在2个像素之内,满足打靶实验的要求.     

正态分布模拟倾斜刃边图像的MTF测量方法

调制传递函数(MTF)是评价光学成像系统信息传递频率特性的一个关键指标,倾斜刃边法是测量MTF的主要方法之一。为了解决传统倾斜刃边法对噪声图像边缘角度估计的不准确性,提出了一种基于正态分布拟合和中值点查寻的MTF测量新算法。通过正态分布函数模拟一幅与实际图像误差最小的模拟图像,利用中值点查寻法对模拟图像处理,求得最佳刃边角度,利用该角度求出实际图像的边缘扩散曲线,再通过费米函数拟合获得边缘扩散函数(ESF)。由实验结果可以看出,在图像噪声较大的情况下,对倾斜刃边角度估计的准确度得到了明显提升,进而也提高了MTF的测量精度。     

基于多约束准则匹配算法的序列图像配准

为了准确、快速地对动态视频序列中的图像进行配准,提出了一种新的基于特征点匹配确定两幅图像间变换关系的方法。首先,根据多约束准则(MR),通过局部信息熵、相似性测度和距离比例不变准则三个约束条件,准确地找到两个点集中的三对匹配点。然后,利用这些点对采用矩阵求解最小二乘法估算两幅图像的仿射变换参数。与相关匹配法相比,MR算法避免了由于图像中运动物体的干扰而产生的伪匹配点对。对于384 pixel×256 pixel的图像,MR算法完成特征点匹配只需2.76 ms,比相关法减少了58%的运算时间。仿射参数在X,Y方向上的估计误差也减小为Δx=0.13,Δy=0.02,远小于1 pixel。基本满足目标检测技术在工程上的速度快、精度高、抗干扰能力强等要求。     

基于ARM的视觉导航AGV图像处理方法研究

为了实现更加简单高效的标识线图像的检测与处理,提出了一种优化的基于ARM的视觉导航AGV标识线图像处理方法。首先对采集到的图像进行灰度化处理并使用Otsu算法对图像进行阈值分割;然后采用优化的中值滤波算法进行图像滤波并使用高效的边缘提取策略获取路径边缘特征;最后采用角度判断的方法剔除错误点并使用最小二乘法拟合成路径的中心线。实验结果表明,该方法有较高的准确率和较好的实时性,可以满足工业生产中的实际需求,适用于基于嵌入式系统开发的视觉导航AGV。     

基于双目光栅重建和纹理映射的缺陷三维测量方法

针对工业检测中对表面缺陷的高精度检测和定位需求,提出了一种缺陷特征重建方法。通过在基于双目光栅投影的三维重构系统上附加纹理相机,实现对于重构点云模型的纹理映射,并结合纹理相机图像中提取到的特征区域,完成表面特征的三维重构。针对待测物体需要进行多视角重建的情况,引入精密转台,利用旋转轴标定方法获取不同旋转位置下纹理图像与点云数据的映射关系,并利用基于距离判据的判断方法实现了对遮挡部分点云的剔除。采用四象限临近点搜索和基于距离加权平均的线性插值方法对纹理图像中像素坐标进行三维测量。实验完成了对于图像中标注缺陷轮廓内像素点的重建,实现了对于表面特征的精确尺寸计算和定位,通过对重建的缺陷尺寸和位置进行计算并与影像测量仪测得结果进行对比,可得本文方案对缺陷三维尺寸和位置的测量误差不超过0.2 mm,且能更准确地计算缺陷面积。     

基于迭代拟合的集装箱角件边缘检测算法

噪声通常是影响集装箱角件图像中低层次语义信息提取精度的重要因素,传统的边缘检测算法通常通过改进滤波器和阈值来消除图像中的物理噪声和环境噪声,但是却无法去除边缘检测后的噪声,为解决这一问题,提出了一种基于迭代拟合的边缘检测算法。首先,对角件图像进行一系列预处理操作获取边缘点集,其次,使用拟合算法处理点集并且得到函数表达式,然后定义偏差值度量并计算,用于衡量目标点集到拟合或者检测结果的偏差,最后,去除定义下距离拟合结果最远的指定数量的点,如此迭代拟合直至评价函数收敛。实验结果与分析表明,该算法可以有效地去除边缘点集中的非真实边缘点,相比于传统的边缘检测算法更能去除特殊噪声,算法具有收敛速度快、准确率较高、灵活性好等特点。     

一种复杂海天背景下红外弱小目标检测算法

红外弱小目标的检测是红外搜索跟踪系统(IRST)中的一项关键技术,常用的目标检测算法存在受海杂波严重、虚警率较高等问题,分析了海天背景下红外图像的背景、小目标的特征,提出了一种海杂波背景下的红外小目标检测算法。首先统计图像的行均值和梯度,用最小二乘法拟合出海天线,然后利用形态学算子抑制图像背景,并采用自适应阈值将图像二值化,最后分析图像的梯度特征,抑制海杂波和云层的干扰。实验结果表明,该方法能精确地提取海天线,稳定地提取海天背景下的弱小目标,虚警率低于5%,目标检测概率超过97%。     

点对称关系耦合距离约束的图像匹配算法

为了解决当前图像匹配算法主要通过图像的区域梯度信息来进行图像匹配,当图像存在光照变化等干扰时,将会使得匹配图像存在较多的错误匹配点以及匹配耗时较长等不足,提出了基于点对称关系耦合距离约束的图像匹配方法。采用FAST算法对图像特征点进行准确、快速的检测,利用拉普拉斯极值模型进一步剔除伪特征点,以提高算法的匹配正确度;对特征点的对称性进行计算,利用点对称关系构造点间距度量模型,以求取特征描述符中的特征向量,输出特征描述符;基于SURF特征点匹配机制,对特征点完成双向匹配约束,完成特征点匹配。对匹配特征点进行欧氏度量,以度量结果的比值以及均值作为依据,构造距离约束模型,利用距离约束模型判别错误匹配点,优化匹配结果。实验结果显示:与当前图像匹配算法相比较,所提算法不仅具有较高的匹配精度以及匹配效率,而且具有较好的鲁棒性能。     

一种基因芯片光学扫描图像倾斜校正方法研究

针对在基因芯片光学扫描时产生的图像倾斜问题,提出了一种基于像素灰度的芯片图像倾斜校正方法。结合基因芯片图像的结构特点,基于行、列方向像素灰度定义芯片图像的校正指标。在角度检测范围内,利用折半搜索方法,基于校正指标来检测芯片图像的校正角度和芯片图像的校正位置。实验结果表明,该方法能有效地检测多类基因芯片图像的倾斜角度,具有较强的鲁棒性和实用性。     

A novel planar calibration method using the iterative correction of control points information

In the camera calibration using translational planar object instead of 3D target, the skew affected by the imprecision in installation causes the object to deviate from the designated position, which produces errors to the world coordinates of control points and reduces the calibration accuracy. In this paper, a skew correction model is established to correct this skew. We note and prove that the skew bias of control point which is close to the fix point is tiny. According to the projection regulation, the pixel distance between two adjacent control points which is close to the fix point in skewed situation is a value extremely approximate to that in the ideal situation. Based on this property, we utilize the pixel distance to assign the image coordinates based on the fix point. The assigned values are then employed to estimate the skew factors, by which the initial camera parameters are optimized and lens distortion in the calibration images are corrected. This process is then repeated until convergence. Experiments based on real images prove that this method is more accurate than other methods without correcting skew. The parameters obtained by our method can be applied to the 3D reconstruction directly and effectively.     

Color image encryption using skew tent map and hyper chaotic system of 6th-order CNN

The paper designs a color image encryption scheme based on skew tent map and hyper chaotic system of 6th-order CNN. The essence of the image encryption is to confuse and diffuse the pixels, the skew tent map is applied to generate the confusion sequence, and the hyper chaotic system of 6th-order CNN is applied to generate the diffusion sequence, for 6 state variables in the system, there are total 120 combinations. For each pixel of the plain image, one combination is chosen to encryption the red, green and blue components, and the combination is determined by one of the state variables. Each pixel is encrypted by the cipher value of the previous pixel and the combination value of the CNN system. Experimental results and security analysis demonstrate that the scheme can achieve good encryption result and larger key space, and can resist common attacks, so the scheme can be applied in secure communication to enhance the security of transmitting image.     

A novel image encryption method based on total shuffling scheme

In this paper, a novel image encryption method based on skew tent chaotic map and permutation-diffusion architecture is proposed. In the proposed method, the P-box is chosen as the same size of plain-image, which shuffles the positions of pixels totally. The keystream generated by skew tent chaotic map is related to the plain-image. Statistical analysis, information entropy analysis, and sensitivity analysis to plaintext and key on the proposed scheme are provided in this paper. It can be seen that this algorithm is efficient and reliable, with high potential to be adopted for network security and secure communications.     

Cryptanalysis on a novel image encryption method based on total shuffling scheme

A novel image encryption method based on a skew tent map is proposed recently. In this paper, some flaws of this algorithm are pointed out and then a chosen plaintext attack against it is presented. Both theoretical analysis and experimental simulation indicate that the plain image can be recovered exactly from the cipher image without the secret key. So it can be seen that this algorithm is not secure enough to be applied in network communication.     

Single-camera panoramic stereo system model based on skew ray tracing

This study integrates panoramic imaging and stereo vision theory to propose an innovative 3D optical positioning method for single-camera panoramic stereo imaging systems. The conventional single-viewpoint theory based on pinhole modeling is replaced by a skew ray tracing methodology to establish the geometrical relationships of object and image. An analytical model of a single-CCD panoramic stereo imaging system is then established by adopting a catadioptric imaging system, a single-CCD stereo panoramic vision system and the presented skew ray camera calibration method.     

A semi-supervised learning approach for detection of phishing webpages

This paper proposes a new phishing webpage detection approach based on a kind of semi-supervised learning method-transductive support vector machine (TSVM). Firstly the features of web image are extracted for complementing the disadvantage of phishing detection only based on document object model (DOM); they include gray histogram, color histogram, and spatial relationship between subgraphs. Then the features of sensitive information are examined by using page analysis based on DOM objects. In contrast to the drawback of support vector machine (SVM) algorithm which simply trains classifier by learning little and poor representative labeled samples, this method introduces the TSVM to train classifier that it takes into account the distribution information implicitly embodied in the large quantity of the unlabeled samples, and have better performance than SVM. The experimental results show that the proposed method not only achieves better classification accuracy, but also has strong applicability as the independent method of phishing detection.     

Note on the Wigner-Yanase-Dyson Skew Information

In this note, releasing the restriction on operators which are observables (self-adjoint), a generalization of the Wigner-Yanase-Dyson skew information is given. We study some properties of the generalization of the Wigner-Yanase-Dyson skew information and related quantities from the operator theory point of view. In particular, an elementary proof of the convexity with the Wigner-Yanase-Dyson skew information is obtained.     

Optimizing head/media electrical characterization in the presence of skew at high track density

The skew angle causes a discrepancy in determining the reader-to-writer offset (RWO) when using different periodical patterns in track profile tests. It also separates the peak overwrite (OW) from the peak high frequency amplitude HFA, (1 T periodical pattern) on corresponding track profiles. Furthermore, higher track density and larger skew angle exacerbate the skew effect and induce more RWO error, thus impacting the parametric performance optimization. Simulation studies are used to interpret the skew effect on the RWO determination and OW cross-track characteristics. Based on experimental investigations and simulation analyses, using the HFA, track profile for deriving the optimal RWO is proposed for spin-stand tests. Actual parametric characterization has proven that the optimal RWO minimized the skew effect and the RWO error, thus improving the parametric performance and reducing the test variation. The method is beneficial and necessary for the high track density characterization.     

Prediction of the performance of an algorithm for the detection of small targets in infrared images

Detection of small targets in infrared (IR) images is important in IR image processing. For the prediction of performance of a detection algorithm, it is necessary to calculate the probability of detection and probability of false alarm. A method is developed to calculate the probabilities in this paper. The detection is divided into two parts: the first part, which is called pre-detection, is to find out candidates for targets in a single frame of an image; and the second part is to localize the target in multiple frames of the image. Under some assumptions, the pre-detection probability, the false detection probability of single frame, detection probability and false alarm probability are derived. The algorithm for the detection of small target in IR image, which is used for the derivation of the probabilities, is contrast threshold detection based on background prediction, and a pipeline filter is used for multiframe image processing. The results show the relationship of the probabilities to the contrast of target to background, SNR, and contrast threshold.