The multitude of airborne point clouds limits the point cloud processing efficiency.Superpoints are grouped based on similar points,which can effectively alleviate the demand for computing resources and improve proces...The multitude of airborne point clouds limits the point cloud processing efficiency.Superpoints are grouped based on similar points,which can effectively alleviate the demand for computing resources and improve processing efficiency.However,existing superpoint segmentation methods focus only on local geometric structures,resulting in inconsistent spectral features of points within a superpoint.Such feature inconsistencies degrade the performance of subsequent tasks.Thus,this study proposes a novel Superpoint Segmentation method that jointly utilizes spatial Geometric and Spectral Information for multispectral point cloud superpoint segmentation(GSI-SS).Specifically,a similarity metric that combines spatial geometry and spectral information is proposed to facilitate the consistency of geometric structures and object attributes within segmented superpoints.Following the formation of the primary superpoints,an intersuperpoint pointexchange mechanism that maximizes feature consistency within the final superpoints is proposed.Experiments are conducted on two real multispectral point cloud datasets,and the proposed method achieved higher recall,precision,F score,and lower global consistency and feature classification errors.The experimental results demonstrate the superiority of the proposed GSI-SS over several state-of-the-art methods.展开更多
煤矿井下视觉同步定位与地图构建SLAM(Simultaneous Localization and Mapping)应用中,光照变化与低纹理场景严重影响特征点的提取和匹配结果,导致位姿估计失败,影响定位精度。提出一种基于改进定向快速旋转二值描述符ORB(Oriented Fast...煤矿井下视觉同步定位与地图构建SLAM(Simultaneous Localization and Mapping)应用中,光照变化与低纹理场景严重影响特征点的提取和匹配结果,导致位姿估计失败,影响定位精度。提出一种基于改进定向快速旋转二值描述符ORB(Oriented Fast and Rotated Brief)-SLAM3算法的煤矿井下移动机器人双目视觉定位算法SL-SLAM。针对光照变化场景,在前端使用光照稳定性的Super-Point特征点提取网络替换原始ORB特征点提取算法,并提出一种特征点网格限定法,有效剔除无效特征点区域,增加位姿估计稳定性。针对低纹理场景,在前端引入稳定的线段检测器LSD(Line Segment Detector)线特征提取算法,并提出一种点线联合算法,按照特征点网格对线特征进行分组,根据特征点的匹配结果进行线特征匹配,降低线特征匹配复杂度,节约位姿估计时间。构建了点特征和线特征的重投影误差模型,在线特征残差模型中添加角度约束,通过点特征和线特征的位姿增量雅可比矩阵建立点线特征重投影误差统一成本函数。局部建图线程使用ORB-SLAM3经典的局部优化方法调整点、线特征和关键帧位姿,并在后端线程中进行回环修正、子图融合和全局捆绑调整BA(Bundle Adjustment)。在EuRoC数据集上的试验结果表明,SL-SLAM的绝对位姿误差APE(Absolute Pose Error)指标优于其他对比算法,并取得了与真值最接近的轨迹预测结果:均方根误差相较于ORB-SLAM3降低了17.3%。在煤矿井下模拟场景中的试验结果表明,SL-SLAM能适应光照变化和低纹理场景,可以满足煤矿井下移动机器人的定位精度和稳定性要求。展开更多
移动机器人在使用单目视觉传感器进行同步定位与建图(simultaneous localization and mapping,SLAM)时,复杂环境中存在大量的光线变化或环境纹理稀疏情况,这是导致其定位不准确的主要因素。因此,文中对ORB-SLAM3系统中前端与定位环节进...移动机器人在使用单目视觉传感器进行同步定位与建图(simultaneous localization and mapping,SLAM)时,复杂环境中存在大量的光线变化或环境纹理稀疏情况,这是导致其定位不准确的主要因素。因此,文中对ORB-SLAM3系统中前端与定位环节进行改进,提升单目视觉移动机器人在复杂环境中的定位精度与鲁棒性。首先,提出区域动态特征概率阈值调整算法对SuperPoint网络进行改进,替换原ORB算法进行图像特征提取,从而获取鲁棒性更强且分布更均匀的视觉特征点;其次,提出共视匹配策略和动态窗口匹配策略,优化了视觉前端的特征匹配与跟踪算法,提升在稀疏纹理场景下的视觉跟踪性能;最后,结合所提改进算法与多传感器信息融合技术,构建了完整的定位系统框架,并在该系统上进行了单目视觉地面移动机器人定位实验。实验结果表明:改进后的算法在EuRoc数据集上的绝对轨迹误差相比ORB-SLAM3降低了8.6%;真实环境中,机器人绝对轨迹误差相比改进前降低了33.59%。展开更多
The superpoints are the sources （or the destinations） that connect with a great deal of destinations （or sources） during a measurement time interval, so detecting the superpoints in real time is very important to ...The superpoints are the sources （or the destinations） that connect with a great deal of destinations （or sources） during a measurement time interval, so detecting the superpoints in real time is very important to network security and management. Previous algorithms are not able to control the usage of the memory and to deliver the desired accuracy, so it is hard to detect the superpoints on a high speed link in real time. In this paper, we propose an adaptive sampling algorithm to detect the superpoints in real time, which uses a flow sample and hold module to reduce the detection of the non-superpoints and to improve the measurement accuracy of the superpoints. We also design a data stream structure to maintain the flow records, which compensates for the flow Hash collisions statistically. An adaptive process based on different sampling probabilities is used to maintain the recorded IP addresses in the limited memory. This algorithm is compared with the other algorithms by analyzing the real network trace data. Experiment results and mathematic analysis show that this algorithm has the advantages of both the limited memory requirement and high measurement accuracy.展开更多
基金supported by the Youth Project of the National Natural Science Foundation of China(Grant No.62201237)the Yunnan Fundamental Research Projects(Grant Nos.202101BE070001-008 and202301AV070003)+1 种基金the Youth Project of the Xingdian Talent Support Plan of Yunnan Province(Grant No.KKRD202203068)the Major Science and Technology Projects in Yunnan Province(Grant No.202202AD080013)。
文摘The multitude of airborne point clouds limits the point cloud processing efficiency.Superpoints are grouped based on similar points,which can effectively alleviate the demand for computing resources and improve processing efficiency.However,existing superpoint segmentation methods focus only on local geometric structures,resulting in inconsistent spectral features of points within a superpoint.Such feature inconsistencies degrade the performance of subsequent tasks.Thus,this study proposes a novel Superpoint Segmentation method that jointly utilizes spatial Geometric and Spectral Information for multispectral point cloud superpoint segmentation(GSI-SS).Specifically,a similarity metric that combines spatial geometry and spectral information is proposed to facilitate the consistency of geometric structures and object attributes within segmented superpoints.Following the formation of the primary superpoints,an intersuperpoint pointexchange mechanism that maximizes feature consistency within the final superpoints is proposed.Experiments are conducted on two real multispectral point cloud datasets,and the proposed method achieved higher recall,precision,F score,and lower global consistency and feature classification errors.The experimental results demonstrate the superiority of the proposed GSI-SS over several state-of-the-art methods.
文摘煤矿井下视觉同步定位与地图构建SLAM(Simultaneous Localization and Mapping)应用中,光照变化与低纹理场景严重影响特征点的提取和匹配结果,导致位姿估计失败,影响定位精度。提出一种基于改进定向快速旋转二值描述符ORB(Oriented Fast and Rotated Brief)-SLAM3算法的煤矿井下移动机器人双目视觉定位算法SL-SLAM。针对光照变化场景,在前端使用光照稳定性的Super-Point特征点提取网络替换原始ORB特征点提取算法,并提出一种特征点网格限定法,有效剔除无效特征点区域,增加位姿估计稳定性。针对低纹理场景,在前端引入稳定的线段检测器LSD(Line Segment Detector)线特征提取算法,并提出一种点线联合算法,按照特征点网格对线特征进行分组,根据特征点的匹配结果进行线特征匹配,降低线特征匹配复杂度,节约位姿估计时间。构建了点特征和线特征的重投影误差模型,在线特征残差模型中添加角度约束,通过点特征和线特征的位姿增量雅可比矩阵建立点线特征重投影误差统一成本函数。局部建图线程使用ORB-SLAM3经典的局部优化方法调整点、线特征和关键帧位姿,并在后端线程中进行回环修正、子图融合和全局捆绑调整BA(Bundle Adjustment)。在EuRoC数据集上的试验结果表明,SL-SLAM的绝对位姿误差APE(Absolute Pose Error)指标优于其他对比算法,并取得了与真值最接近的轨迹预测结果:均方根误差相较于ORB-SLAM3降低了17.3%。在煤矿井下模拟场景中的试验结果表明,SL-SLAM能适应光照变化和低纹理场景,可以满足煤矿井下移动机器人的定位精度和稳定性要求。
文摘移动机器人在使用单目视觉传感器进行同步定位与建图(simultaneous localization and mapping,SLAM)时,复杂环境中存在大量的光线变化或环境纹理稀疏情况,这是导致其定位不准确的主要因素。因此,文中对ORB-SLAM3系统中前端与定位环节进行改进,提升单目视觉移动机器人在复杂环境中的定位精度与鲁棒性。首先,提出区域动态特征概率阈值调整算法对SuperPoint网络进行改进,替换原ORB算法进行图像特征提取,从而获取鲁棒性更强且分布更均匀的视觉特征点;其次,提出共视匹配策略和动态窗口匹配策略,优化了视觉前端的特征匹配与跟踪算法,提升在稀疏纹理场景下的视觉跟踪性能;最后,结合所提改进算法与多传感器信息融合技术,构建了完整的定位系统框架,并在该系统上进行了单目视觉地面移动机器人定位实验。实验结果表明:改进后的算法在EuRoc数据集上的绝对轨迹误差相比ORB-SLAM3降低了8.6%;真实环境中,机器人绝对轨迹误差相比改进前降低了33.59%。
基金the National Basic Research Program of China (Grant No. 2003cb314804)
文摘The superpoints are the sources （or the destinations） that connect with a great deal of destinations （or sources） during a measurement time interval, so detecting the superpoints in real time is very important to network security and management. Previous algorithms are not able to control the usage of the memory and to deliver the desired accuracy, so it is hard to detect the superpoints on a high speed link in real time. In this paper, we propose an adaptive sampling algorithm to detect the superpoints in real time, which uses a flow sample and hold module to reduce the detection of the non-superpoints and to improve the measurement accuracy of the superpoints. We also design a data stream structure to maintain the flow records, which compensates for the flow Hash collisions statistically. An adaptive process based on different sampling probabilities is used to maintain the recorded IP addresses in the limited memory. This algorithm is compared with the other algorithms by analyzing the real network trace data. Experiment results and mathematic analysis show that this algorithm has the advantages of both the limited memory requirement and high measurement accuracy.
