核双样检验的连杆摆动光流轨迹混叠去除及补偿  

Removal and compensation of mechanical rod swing optical flow trajectory aliasing based on kernel double sample test

作　　者：郑思凡 陈平平[1,3] 苏凯雄[1,3] 吴永春 林志坚[1] Zheng Sifan;Chen Pingping;Su Kaixiong;Wu Yongchun;Lin Zhijian(College of Physics and Information Engineering,Fuzhou University,Fuzhou 350116,China;Intelligent Manufacturing Engineering Institute,Liming Vocational University,Quanzhou 362000,China;Digital TV Intelligent Technology National and Local Joint Engineering Research Center,Fuzhou 350116,China)

机构地区：[1]福州大学物理与信息工程学院,福州350116 [2]黎明职业大学智能制造工程学院,泉州362000 [3]数字电视智能化技术国家地方联合工程研究中心,福州350116

出　　处：《仪器仪表学报》2023年第7期182-193,共12页Chinese Journal of Scientific Instrument

基　　金：福建省软件业技术创新重点攻关及产业化项目(350116C025-2023-7);福建省政府招标采购交通执法信息系统运维服务项目(3500FRGK2017008-1)资助。

摘　　要：在连杆摆动视觉测量中,针对高速曝光产生的背景噪点对前景光流轨迹混叠及阻断导致配准误差,本文提出了一种新的视频滤波与补偿算法。该算法以背景噪声与前景摆杆光流轨迹不同的运动统计特性作为先验模型,通过核双样假设检验检测光流轨迹速度突变,剪除背景噪点轨迹片段实现去噪。为实现轨迹补偿,首先引入机械连杆铰接点为参照物的相对光流采集方式,将各帧摆杆铰接点配准翘曲至第1帧铰接点位置,将连杆摆动轨迹从复合高阶摆线中分离形成低阶理想圆弧。其次采用Pratt拟合翘曲后的摆杆轨迹的圆心与半径,将轨迹聚类为不同半径的弧状轨迹群。最后,将弧状轨迹群的x-y坐标ν-SVR回归作为几何约束,结合x-t动力学回归半监督学习出完整长度的轨迹。在刮刷总成摆角及共轭凸轮的针床推程位移比较的测量实验表明,该算法比传统VBM3D,MeshFlow等算法准确度可提高3.26%,运算复杂度降低2阶,在机械旋转运动视觉故障诊断及机械仪表数字化采集等方面具有广阔应用前景。In the visual measurement of rod′s angular velocity,the background noise caused by short exposure time leads to serious registration error by mixing and blocking optical flow trajectory.To address this issue,the article proposes a trajectory compensation and noise clipping algorithm.The algorithm takes the difference motion statistical characteristics between background noise and rod′s optical flow trajectory as prior model,and cuts out the background noise trajectory segments to remove the trajectory aliasing by detecting the speed mutation points through the kernel double sample hypothesis test.By introducing a new optical flow acquisition method which takes the rod hinge fulcrum as the reference,the rod′s hinge points of each frame are registered and warped to the position of the first frame to separate the rod′s trajectory from the total composite high-order cycloid to form a low order ideal arc as compensation prior.Secondly,the trajectory is clustered into arc shaped trajectory groups with different radii which fitted by Pratt.Thirdly,the full length trajectory is semi-supervised learned from theν-SVR regression of arc trajectory groups which acts as a geometric constraint and from x-t dynamic regression to realize compensation.The comparative measurement experiment of the brush assembly angular velocity and displacement of the needle bed conjugate cam shows that this algorithm can improve the accuracy by 3.26%compared to traditional algorithms such as VBM3D and MeshFlow.The computational complexity is reduced by 2 orders.It has broad application prospects in visual fault diagnosis of mechanical rotation motion and digital acquisition of mechanical instruments.

关 键 词：连杆曲线图谱 BEBLID描述子 最大均值差异 RKHS再生核 密度功率散度 核双样检验 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] TH691[自动化与计算机技术—控制科学与工程]