针对ORB(oriented FAST and rotated BRIEF)算法中存在匹配精确率低的问题,提出了一种基于LK(Lucas-Kanade)光流改进的ORB图像匹配方法。首先对待处理的图像进行直方图均衡化,然后在Oriented FAST特征点检测的同时用LK光流对其进行跟踪...针对ORB(oriented FAST and rotated BRIEF)算法中存在匹配精确率低的问题,提出了一种基于LK(Lucas-Kanade)光流改进的ORB图像匹配方法。首先对待处理的图像进行直方图均衡化,然后在Oriented FAST特征点检测的同时用LK光流对其进行跟踪,并将跟踪的特征点进行Rotated BRIEF描述,最后在特征匹配筛选环节利用RANSAC(Random Sampling Consistency)算法进行误匹配的剔除。实验结果表明,改进算法在公开数据集中的平均匹配精度为90.9%,平均特征匹配及误匹配的剔除共耗时为18ms,与原始ORB算法相比,在时间基本一致的前提下,有效的提高了匹配的精度。展开更多
Tracking registration is a key issue in augmented reality applications,particularly where there are no artificial identifier placed manually.In this paper,an efficient markerless tracking registration algorithm which ...Tracking registration is a key issue in augmented reality applications,particularly where there are no artificial identifier placed manually.In this paper,an efficient markerless tracking registration algorithm which combines the detector and the tracker is presented for the augmented reality system.We capture the target images in real scenes as template images,use the random ferns classifier for target detection and solve the problem of reinitialization after tracking registration failures due to changes in ambient lighting or occlusion of targets.Once the target has been successfully detected,the pyramid Lucas-Kanade(LK)optical flow tracker is used to track the detected target in real time to solve the problem of slow speed.The least median of squares(LMedS)method is used to adaptively calculate the homography matrix,and then the three-dimensional pose is estimated and the virtual object is rendered and registered.Experimental results demonstrate that the algorithm is more accurate,faster and more robust.展开更多
Lucas-Kanade(LK) algorithm, usually used in optical flow filed, has recently received increasing attention from PIV community due to its advanced calculation efficiency by GPU acceleration. Although applications of th...Lucas-Kanade(LK) algorithm, usually used in optical flow filed, has recently received increasing attention from PIV community due to its advanced calculation efficiency by GPU acceleration. Although applications of this algorithm are continuously emerging,a systematic performance evaluation is still lacking. This forms the primary aim of the present work. Three warping schemes in the family of LK algorithm: forward/inverse/symmetric warping, are evaluated in a prototype flow of a hierarchy of multiple two-dimensional vortices. Second-order Newton descent is also considered here. The accuracy & efficiency of all these LK variants are investigated under a large domain of various influential parameters. It is found that the constant displacement constraint, which is a necessary building block for GPU acceleration, is the most critical issue in affecting LK algorithm's accuracy, which can be somehow ameliorated by using second-order Newton descent. Moreover, symmetric warping outbids the other two warping schemes in accuracy level, robustness to noise, convergence speed and tolerance to displacement gradient, and might be the first choice when applying LK algorithm to PIV measurement.展开更多
针对同步定位与地图建立(simultaneous localization and mapping,SLAM)算法在动态环境下存在位姿估计和地图构建误差较大的问题,提出一种光流语义分割方法用于增加动态场景下的可运行性。将ORB-SLAM2系统与YOLOv5模型结合,对传入图像...针对同步定位与地图建立(simultaneous localization and mapping,SLAM)算法在动态环境下存在位姿估计和地图构建误差较大的问题,提出一种光流语义分割方法用于增加动态场景下的可运行性。将ORB-SLAM2系统与YOLOv5模型结合,对传入图像提取特征点的同时将YOLOv5网络模型语义分割后的物体分为高、中、低动态物体。利用运动一致性检测算法,对三种检测物体动态阈值判断,辨别其是否需要剔除特征点,增加ORB-SLAM2算法在动态环境下的运行精度。为加快系统运行速度,用LK光流法加快普通帧与普通帧之间的匹配,其原理为使用LK光流匹配特征点代替ORB特征点匹配,大大的缩小运行时间,同时运行误差变化不大。实验在TUM数据集下测试,平均每一帧提取2000个特征点,在增加LK光流后缩短0.01 s以上,若在900帧数据集下,可缩短9 s.其绝对轨迹误差对比于ORB-SLAM2和DS-SLAM平均提升在95%与30%以上,证明了算法在动态场景下良好的运行精度与鲁棒性。展开更多
文摘针对ORB(oriented FAST and rotated BRIEF)算法中存在匹配精确率低的问题,提出了一种基于LK(Lucas-Kanade)光流改进的ORB图像匹配方法。首先对待处理的图像进行直方图均衡化,然后在Oriented FAST特征点检测的同时用LK光流对其进行跟踪,并将跟踪的特征点进行Rotated BRIEF描述,最后在特征匹配筛选环节利用RANSAC(Random Sampling Consistency)算法进行误匹配的剔除。实验结果表明,改进算法在公开数据集中的平均匹配精度为90.9%,平均特征匹配及误匹配的剔除共耗时为18ms,与原始ORB算法相比,在时间基本一致的前提下,有效的提高了匹配的精度。
基金supported by National Natural Science Foundation of China(No.61125101).
文摘Tracking registration is a key issue in augmented reality applications,particularly where there are no artificial identifier placed manually.In this paper,an efficient markerless tracking registration algorithm which combines the detector and the tracker is presented for the augmented reality system.We capture the target images in real scenes as template images,use the random ferns classifier for target detection and solve the problem of reinitialization after tracking registration failures due to changes in ambient lighting or occlusion of targets.Once the target has been successfully detected,the pyramid Lucas-Kanade(LK)optical flow tracker is used to track the detected target in real time to solve the problem of slow speed.The least median of squares(LMedS)method is used to adaptively calculate the homography matrix,and then the three-dimensional pose is estimated and the virtual object is rendered and registered.Experimental results demonstrate that the algorithm is more accurate,faster and more robust.
基金supported by the National Natural Science Foundation of China(Grant Nos.11372001 and 11490552)
文摘Lucas-Kanade(LK) algorithm, usually used in optical flow filed, has recently received increasing attention from PIV community due to its advanced calculation efficiency by GPU acceleration. Although applications of this algorithm are continuously emerging,a systematic performance evaluation is still lacking. This forms the primary aim of the present work. Three warping schemes in the family of LK algorithm: forward/inverse/symmetric warping, are evaluated in a prototype flow of a hierarchy of multiple two-dimensional vortices. Second-order Newton descent is also considered here. The accuracy & efficiency of all these LK variants are investigated under a large domain of various influential parameters. It is found that the constant displacement constraint, which is a necessary building block for GPU acceleration, is the most critical issue in affecting LK algorithm's accuracy, which can be somehow ameliorated by using second-order Newton descent. Moreover, symmetric warping outbids the other two warping schemes in accuracy level, robustness to noise, convergence speed and tolerance to displacement gradient, and might be the first choice when applying LK algorithm to PIV measurement.
