基于边缘角点的SIFT图像配准算法

SIFT由特征提取,特征描述符描述和特征匹配3部分构成,该算子特征提取数目庞大,建立特征描述符运算量高,导致算法效率低。提出了一种SEC(SIFT-Edge-Corner)算法,在图像尺度空间提取角点代替SIFT特征点,并根据角点是边缘曲率极值理论,预先采用Canny算子得到高斯边缘图像金字塔,再提取角点并进行尺度选择。实验结果表明:该算法在保障高准确率的前提下大幅度提高特征提取效率。     

基于边缘角点的SIFT图像配准算法

SIFT由特征提取,特征描述符描述和特征匹配3部分构成,该算子特征提取数目庞大,建立特征描述符运算量高,导致算法效率低.提出了一种SEC (SIFT-Edge-Corner)算法,在图像尺度空间提取角点代替SIFT特征点,并根据角点是边缘曲率极值理论,预先采用Canny算子得到高斯边缘图像金字塔,再提取角点并进行尺度选择.实验结果表明:该算法在保障高准确率的前提下大幅度提高特征提取效率.     

一种半自动的基于图像建模系统

针对目前各种类型航海模拟器的虚拟建模效率较低与真实感较差等问题,设计与实现了一种基于多幅图像及少量用户交互恢复物体三维几何模型的系统.本系统采用稳定性好的改进Har-ris算子进行角点检测,基于角点空间特征及交叉相关实现角点的初始匹配;采用基于极线约束的角点匹配快速算法完成角点的精匹配,同时引入迭代和人工辅助修正的方法解决匹配算法的退化情况.在角点匹配结果基础上进行仿射重建,通过升级矩阵实现度量重建,在OpenGL环境下绘制出物体的三维模型框架.从原图像中抽取纹理并采用多视角图像融合的方法进行纹理映射,得到"照片级"真实感的三维模型.     

基于计算机视觉的平面角度测量方法

为解决制造、地质、控制等领域中角度的快速测量问题,提出了一种基于计算机视觉原理的非接触式角度测量方法.测量时仅需用已知内部参数的相机,在空间任意位置对待测角进行1次拍照即可.在所得图像上,利用直线检测和最小二乘法,得到待测角各边1组平行线或1段均分线段的图像方程;再通过计算平行线图像的交点或利用交比的调和性,确定对应边上灭点的坐标;根据相机成像方程,即可由灭点坐标和已知的相机内部参数计算出待测角的角度.实验证明,用此方法,即使在较远距离处测量,其相对误差也能控制在2%以内.     

基于遥感图像信息特征的单调递增SSDA算法

图像配准就是将不同时间、不同传感器或不同条件下对同一景物获取的两幅或多幅图像,进行比较找到该组图像中的共有景物,或是根据已知模式到另一幅图中寻找相应的模式。利用遥感图像进行目标监测需要进行图像配准处理。匹配算法如何达到高精度、高匹配正确率和实时性成为人们追求的目标。文章在传统匹配算法的基础上,提出两点改进:一是通过PCA-圆形SIFT算法提取图像特征角点,降低维数,优化计算;二是利用图像角点作为单调递增阈值序列SSDA算法匹配的基本像素点,利用遥感图像信息特征降低匹配计算量。最后进行实验仿真,结果表明,改进后的算法使得配准速度进一步提高,能够满足遥感图像配准实时性的要求。     

Vuforia的AR识别图像目标设计探析

文章基于计算机视觉中的角点检测算法,以Vuforia AR SDK的目标星级评定系统为例,从图形细节与角点分布,图案内容重复性与角点尺度不变性、旋转不变性,图片明暗反差阈值与角点数量三个方面探析作为AR识别物的图像目标的设计要点,以期为从事AR应用开发的美术设计人员提供通俗易懂的参考,为现今AR图像目标在美术设计中面临的系统化、规范化文献缺失问题,提供一个探析方向。     

基于HALCON的PCB图像拼接方法研究

以印刷电路板(PCB)贴装生产中的视觉检测为背景研究图像拼接方法.方法包括图像预处理、特征点提取和拼接图像三个步骤,并且在拼接图像步骤中使用了三种技术.利用Halcon软件以两幅PCB图像为目标进行了三个实验测试,并用Do EM方法对实验结果进行评价,评价结果分别为:0.9260(实验1),0.8106(实验2),0.7635(实验3).实验1对应的图像拼接方法能以较少的匹配点数得到最优的匹配结果,可用于PCB贴装质量的视觉检测.     

基于点特征的图像数据配准方法研究

数据配准是数据融合过程中最重要的一个环节,通过感兴趣点检测算子提取点特征,利用点特征中的控制点或特殊点来定义松弛算法中的基本点对,并对松弛算法中点特征的匹配度进行改进,从而完成基于点特征的图像配准过程.实验表明,在仅存点特征位置偏移和存在比例、旋转与平移变化2种情况下的点特征匹配中,文中所提出的匹配算法有很高的匹配正确率.     

基于高斯金字塔的视觉里程计算法研究

针对视觉里程计常用角点提取算法因角点分布不均匀而导致运动信息偏差较大的问题,提出一种基于高斯金字塔的角点提取算法。该算法在角点提取过程中先采用高斯金字塔算法对图片进行尺度压缩,纹理丰富区域压缩纹理,纹理稀疏区域聚集纹理,得到小尺度顶图;然后采用Shi-Tomasi算法提取小尺度顶图角点特征以实现角点粗定位,最后将粗定位信息映射回细节丰富的原图进行角点精准定位,得到图片特征信息。最后,利用金字塔LK光流法追踪角点,根据相机对极几何约束模型恢复运动信息。论文采用KITTI数据集,与原Shi-Tomasi算法、Harris算法、Fast算法进行了对比实验,结果表明本算法可有效改善角点分布均匀性,提高视觉里程计运动信息恢复的精度。     

基于B样条函数的改进Harris角点检测算法

针对Harris算法中存在的角点定位不精确,高斯平滑窗口大小不易选择等问题,提出了一种改进算法,利用B样条函数来代替算法中的高斯平滑函数,可避免高斯窗口的选择,并提高角点的定位精度.通过实验对改进算法的性能进行了分析,证明改进算法能够有效提高角点定位精度.     

利用角点信息的嘴唇轮廓提取

提出了利用角点信息提取嘴唇特征的方法.在可变形模板的框架下,利用以前方法中没有用到的角点信息,提出了基于角点信息的能量函数.该能量函数能够更为精确地反映嘴角点和牙齿区域的特征,跟踪嘴唇的变形.实验结果表明,该方法能更鲁棒地提取嘴唇轮廓特征.     

基于轻量化网络和注意力机制的智能车快速目标识别方法

为提高智能车在真实环境中的实时检测能力，改善复杂环境下检测效果不佳的问题，本文提出一种基于轻量化网络和注意力机制的智能车快速目标识别方法。首先，为了减少网络计算参数和提升目标识别算法的推理速度，提出利用GhostNet加速YOLOv4的特征提取；其次，为了提高复杂场景下对道路目标的识别精度，在GhostNet和特征金字塔部分添加结合软阈值化改进的注意力模块；最后，为了验证本文提出方法的有效性，选取Pascal VOC、KITTI公开数据集和自制城市道路数据集进行实验对比。与其他目标检测算法在精度和速度上进行比较，结果证明，本文方法在平均检测精度提升1.7%的情况下，模型参数量降低到原来的18.7%，检测速度提升了 66%，检测速度和精度均优于其他算法，可满足智能车的实时感知需求。     

Five-Axis Interpolation of Continuous Short Linear Trajectories for 3[PP]S-XY Hybrid Mechanism by Dual Bezier Blending

A novel five-axis real-time interpolation algorithm for 3[PP]S-XY hybrid mechanism is proposed in this paper. In the algorithm, the five-axis tool path for controlling this hybrid mechanism is separated into two sub-paths. One sub-path describes the movement of 3[PP]S parallel kinematic mechanism module, and the other one describes the movement of XY platform. A pair of cubic Bezier curves is employed to smooth the corners in those two sub-paths. Based on the homogenous Jacobian matrix of 3[PP]S mechanism, a relationship between the position errors of every driving joint in hybrid mechanism and the position deviation of the tool tip center point at the moving platform is established. This relationship is used to estimate the approximation error for the corners smoothing according to the accuracy requirement of tool tip center in interpolation. Due to the high computational efficiency of this corner smoothing method, it is integrated into the look-ahead module of computer numerical control(CNC) system to perform online tool path smoothing. By performing the speed planning based on a floating window scheme, a jerk limited S-shape speed profile can be generated efficiently. On this basis, a realtime look-ahead scheme, which is comprised of path-smoothing and feedrate scheduling, is developed to acquire a speed profile with smooth acceleration. A monotonic cubic spline is employed for synchronization between those two smoothed sub-paths in tool path interpolation. This interpolation algorithm has been integrated into our own developed CNC system to control a 3PRS-XY experimental instrument(P, R and S standing for prismatic,revolute and spherical, respectively). A club shaped trajectory is adopted to verify the smoothness and efficiency of the five-axis interpolator for hybrid mechanism control.     

Real-time follow-up head tracking in dynamic complex environments

In the modes of both object motion and camera motion, an enhanced Camshift algorithm, which is based on suppressing similar color features of background and on joint color probability density distribution image, is proposed to real-time track head in dynamic complex environment. The system consists of face detection module, head tracking module and camera control module. When tracking fails, a self-recovery mechanism is introduced. At first the Adaboost face detector based on Haar-like features is implemented to find frontal faces, the false positive is filtered according to the skin color criterion, and the true face is used to initialize the tracking module. In hue saturation value (HSV) colorspace, the hue-saturation (H-S) histogram of face skin and the saturation-value (S-V) histogram of hair are built to produce the joint color probability density distribution image, and this is intended to realize the head tracking with arbitrary pose. During tracking, region of interest (ROI) is introduced, and the color probability density distribution of a specified background area outside the ROI is learned, similar color features in the head are suppressed according to the learning result. The background suppression step is intended to resolve the problem that the tracker maybe fails when the head is distracted by backgrounds having similar colors with the head. A closed loop control model based on speed regulation is applied to drive an active camera to center the head. Once tracking drift or failure is detected, the system stops tracking and returns to the face detection module. Our experimental results show that the presented system is well suitable for tracking head with arbitrary pose in dynamic complex environments, also the active camera can track moving head smoothly and stably. The system is computationally efficient and can run in real-time completely.     

电动汽车无线供电系统电能发射线圈设计与切换控制

针对采用级联式发射线圈的电动汽车无线供电系统中线圈切换时存在的互感急剧下降及汽车位置检测困难的问题,提出了一种对嵌式电能发射线圈,并根据互感稳恒原则及其计算方法,给出了对嵌式电能发射线圈主要参数的设计方法,提出了一种双线圈式车体位置检测传感器,给出了传感器的尺寸参数设计方法及电能传输系统对传感器的干扰抑制方法,阐述了级联式发射线圈的切换控制策略.基于Ansoft Maxwell平台、Matlab/Simulink平台和电动汽车ICPT无线供电系统实验平台分别对研究成果进行了仿真分析和实验验证,结果表明:实验实测互感值波动率约为8%,车载拾取电压波动率约为10%,对嵌式能量发射线圈能够有效地缓解ICPT无线供电电动车在切换过程中的互感下降问题;双线圈式车体位置检测系统能够有效地在40 kHz能量通道电磁场的干扰中拾取位置信号,表明该位置检测方案及切换控制策略的可行性.      

信息共享对树形供应链中的实体影响研究

综合考虑影响供应链性能的参数,建立了基于WEB模式的供应链模拟平台。利用此平台,研究较复杂的三阶树形供应链结构下,信息共享对供应链中实体的影响。分别在不共享与共享库存信息下进行实验,实验数据表明,共享库存信息对上游实体影响比对下游实体的影响更大,共享库存信息可使上游实体获得更大的利益。     

Five-axis interpolation of continuous short linear trajectories for 3[PP]S-<Emphasis Type="Italic">XY</Emphasis> hybrid mechanism by dual Bezier blending

A novel five-axis real-time interpolation algorithm for 3[PP]S-XY hybrid mechanism is proposed in this paper. In the algorithm, the five-axis tool path for controlling this hybrid mechanism is separated into two sub-paths. One sub-path describes the movement of 3[PP]S parallel kinematic mechanism module, and the other one describes the movement of XY platform. A pair of cubic Bezier curves is employed to smooth the corners in those two sub-paths. Based on the homogenous Jacobian matrix of 3[PP]S mechanism, a relationship between the position errors of every driving joint in hybrid mechanism and the position deviation of the tool tip center point at the moving platform is established. This relationship is used to estimate the approximation error for the corners smoothing according to the accuracy requirement of tool tip center in interpolation. Due to the high computational efficiency of this corner smoothing method, it is integrated into the look-ahead module of computer numerical control (CNC) system to perform online tool path smoothing. By performing the speed planning based on a floating window scheme, a jerk limited S-shape speed profile can be generated efficiently. On this basis, a realtime look-ahead scheme, which is comprised of path-smoothing and feedrate scheduling, is developed to acquire a speed profile with smooth acceleration. A monotonic cubic spline is employed for synchronization between those two smoothed sub-paths in tool path interpolation. This interpolation algorithm has been integrated into our own developed CNC system to control a 3PRS-XY experimental instrument (P, R and S standing for prismatic, revolute and spherical, respectively). A club shaped trajectory is adopted to verify the smoothness and efficiency of the five-axis interpolator for hybrid mechanism control.     

液力自动变速器参数模糊匹配

把发动机与液力变矩器组合为一种新的动力源,建立相关的数学模型,以燃料经济性为目标,以动力性为约束奈件,用模糊匹配的方法对车辆传动参数及结构参数进行模糊匹配计算,可获得满意的匹配效果。