为降低视觉设备感知航行环境时,水面光照反射对船舶位姿估计和环境地图重构的影响,在HSV(hue,saturation,value)颜色空间下,采用K均值聚类算法对近岸航行环境图像进行聚类分割处理。改进快速特征点提取和描述算法(oriented FAST and rot...为降低视觉设备感知航行环境时,水面光照反射对船舶位姿估计和环境地图重构的影响,在HSV(hue,saturation,value)颜色空间下,采用K均值聚类算法对近岸航行环境图像进行聚类分割处理。改进快速特征点提取和描述算法(oriented FAST and rotated BRIEF,ORB)来提高即时定位与地图构建(simultaneous localization and mapping,SLAM)效率,缩短特征点匹配时间,改善对外界环境的感知效果并提升船舶自身位姿估计精度。采用2020年南宁海事局执法船进港和靠泊期间由单目相机拍摄的视频数据进行实例验证。结果表明,提出的算法比传统SLAM算法的运行耗时更少,与传统定位设备输出轨迹的偏差较小,可为船舶全面立体感知海上航行环境提供研究基础。展开更多
In order to develop an automated segmentation system for Computed Tomography （CT） brain images, a new approach which consists of several unsupervised segmcotation techniques was introduced. The system segments the C...In order to develop an automated segmentation system for Computed Tomography （CT） brain images, a new approach which consists of several unsupervised segmcotation techniques was introduced. The system segments the CT brain images into three partitions, i. e., abnormalities, cerebrospinal fluid （CSF）, and brain matter. Our approach consists of two phase-segmentation methods. In the first phase segmentation, k-means and fuzzy cmeans （FCM） methods were implemented to segment and transform the images into the binary images. Based on the connected component in binary images, a decision tree was employed for the annotation of normal or abnormal regions. In the second phase segmentation, the modified FCM with population-diameter independent （PDI） segmentation was applied to segment the images into CSF and brain matter. The experimental results have shown that our proposed system is feasible and yield satisfactory results.展开更多
One of the most important problems of clustering is to define the number of classes. In fact, it is not easy to find an appropriate method to measure whether the cluster configuration is acceptable or not. In this pap...One of the most important problems of clustering is to define the number of classes. In fact, it is not easy to find an appropriate method to measure whether the cluster configuration is acceptable or not. In this paper we propose a possible and non-automatic solution considering different criteria of clustering and comparing their results. In this way robust structures of an analyzed dataset can be often caught （or established） and an optimal cluster configuration, which presents a meaningful association, may be defined. In particular, we also focus on the variables which may be used in cluster analysis. In fact, variables which contain little clustering information can cause misleading and not-robustness results. Therefore, three algorithms are employed in this study： K-means partitioning methods, Partitioning Around Medoids （PAM） and the Heuristic Identification of Noisy Variables （HINoV）. The results are compared with robust methods ones.展开更多
文摘为降低视觉设备感知航行环境时,水面光照反射对船舶位姿估计和环境地图重构的影响,在HSV(hue,saturation,value)颜色空间下,采用K均值聚类算法对近岸航行环境图像进行聚类分割处理。改进快速特征点提取和描述算法(oriented FAST and rotated BRIEF,ORB)来提高即时定位与地图构建(simultaneous localization and mapping,SLAM)效率,缩短特征点匹配时间,改善对外界环境的感知效果并提升船舶自身位姿估计精度。采用2020年南宁海事局执法船进港和靠泊期间由单目相机拍摄的视频数据进行实例验证。结果表明,提出的算法比传统SLAM算法的运行耗时更少,与传统定位设备输出轨迹的偏差较小,可为船舶全面立体感知海上航行环境提供研究基础。
文摘In order to develop an automated segmentation system for Computed Tomography （CT） brain images, a new approach which consists of several unsupervised segmcotation techniques was introduced. The system segments the CT brain images into three partitions, i. e., abnormalities, cerebrospinal fluid （CSF）, and brain matter. Our approach consists of two phase-segmentation methods. In the first phase segmentation, k-means and fuzzy cmeans （FCM） methods were implemented to segment and transform the images into the binary images. Based on the connected component in binary images, a decision tree was employed for the annotation of normal or abnormal regions. In the second phase segmentation, the modified FCM with population-diameter independent （PDI） segmentation was applied to segment the images into CSF and brain matter. The experimental results have shown that our proposed system is feasible and yield satisfactory results.
文摘One of the most important problems of clustering is to define the number of classes. In fact, it is not easy to find an appropriate method to measure whether the cluster configuration is acceptable or not. In this paper we propose a possible and non-automatic solution considering different criteria of clustering and comparing their results. In this way robust structures of an analyzed dataset can be often caught （or established） and an optimal cluster configuration, which presents a meaningful association, may be defined. In particular, we also focus on the variables which may be used in cluster analysis. In fact, variables which contain little clustering information can cause misleading and not-robustness results. Therefore, three algorithms are employed in this study： K-means partitioning methods, Partitioning Around Medoids （PAM） and the Heuristic Identification of Noisy Variables （HINoV）. The results are compared with robust methods ones.
