A two-stage state recognition method is proposed for asynchronous SSVEP(steady-state visual evoked potential) based brain-computer interface(SBCI) system.The two-stage method is composed of the idle state(IS) detectio...A two-stage state recognition method is proposed for asynchronous SSVEP(steady-state visual evoked potential) based brain-computer interface(SBCI) system.The two-stage method is composed of the idle state(IS) detection and control state(CS) discrimination modules.Based on blind source separation and continuous wavelet transform techniques,the proposed method integrates functions of multi-electrode spatial filtering and feature extraction.In IS detection module,a method using the ensemble IS feature is proposed.In CS discrimination module,the ensemble CS feature is designed as feature vector for control intent classification.Further,performance comparisons are investigated among our IS detection module and other existing ones.Also the experimental results validate the satisfactory performance of our CS discrimination module.展开更多
A framework is presented for robustly estimating the location of a mobile robot in urban areas based on images extracted from a monocular onboard camera, given a 2D map with building outlines with neither 3D geometric...A framework is presented for robustly estimating the location of a mobile robot in urban areas based on images extracted from a monocular onboard camera, given a 2D map with building outlines with neither 3D geometric information nor appearance data. The proposed method firstly reconstructs a set of vertical planes by sampling and clustering vertical lines from the image with random sample consensus (RANSAC), using the derived 1D homographies to inform the planar model. Then, an optimal autonomous localization algorithm based on the 2D building boundary map is proposed. The physical experiments are carried out to validate the robustness and accuracy of our localization approach.展开更多
在SLAM领域中,为了克服稀疏特征地图不能提供详尽环境信息的缺点,从观测信息的物理意义出发,提出了全局观测地图模型.其基本思想是在稀疏特征地图中嵌入全局密集地图信息,采用位移准则、特征准则和传感器量程准则提取必要的观测信息,然...在SLAM领域中,为了克服稀疏特征地图不能提供详尽环境信息的缺点,从观测信息的物理意义出发,提出了全局观测地图模型.其基本思想是在稀疏特征地图中嵌入全局密集地图信息,采用位移准则、特征准则和传感器量程准则提取必要的观测信息,然后对观测信息进行去噪、转换,接着根据观测信息的物理意义和机器人位姿估计的不确定性获取环境的全局密集地图,可视化后得到环境的二值地图、灰度地图或颜色地图.将全局观测地图模型与EKF-SLAM算法相结合,提出了GOE-SLAM算法,采用Car Park Dataset对GOE-SLAM进行了实验验证,结果表明GOE-SLAM生成了可信的密集地图,并且GOE-SLAM的计算复杂度与EKF-SLAM相当.展开更多
基金National Natural Science Foundation of China(90820305,60775040)
文摘A two-stage state recognition method is proposed for asynchronous SSVEP(steady-state visual evoked potential) based brain-computer interface(SBCI) system.The two-stage method is composed of the idle state(IS) detection and control state(CS) discrimination modules.Based on blind source separation and continuous wavelet transform techniques,the proposed method integrates functions of multi-electrode spatial filtering and feature extraction.In IS detection module,a method using the ensemble IS feature is proposed.In CS discrimination module,the ensemble CS feature is designed as feature vector for control intent classification.Further,performance comparisons are investigated among our IS detection module and other existing ones.Also the experimental results validate the satisfactory performance of our CS discrimination module.
基金National Nature Science Foundation of China(60905061)
文摘A framework is presented for robustly estimating the location of a mobile robot in urban areas based on images extracted from a monocular onboard camera, given a 2D map with building outlines with neither 3D geometric information nor appearance data. The proposed method firstly reconstructs a set of vertical planes by sampling and clustering vertical lines from the image with random sample consensus (RANSAC), using the derived 1D homographies to inform the planar model. Then, an optimal autonomous localization algorithm based on the 2D building boundary map is proposed. The physical experiments are carried out to validate the robustness and accuracy of our localization approach.
文摘在SLAM领域中,为了克服稀疏特征地图不能提供详尽环境信息的缺点,从观测信息的物理意义出发,提出了全局观测地图模型.其基本思想是在稀疏特征地图中嵌入全局密集地图信息,采用位移准则、特征准则和传感器量程准则提取必要的观测信息,然后对观测信息进行去噪、转换,接着根据观测信息的物理意义和机器人位姿估计的不确定性获取环境的全局密集地图,可视化后得到环境的二值地图、灰度地图或颜色地图.将全局观测地图模型与EKF-SLAM算法相结合,提出了GOE-SLAM算法,采用Car Park Dataset对GOE-SLAM进行了实验验证,结果表明GOE-SLAM生成了可信的密集地图,并且GOE-SLAM的计算复杂度与EKF-SLAM相当.