A learning-based control approach is presented for force servoing of a robot with vision in an unknown environment. Firstly, mapping relationships between image features of the servoing object and the joint angles of ...A learning-based control approach is presented for force servoing of a robot with vision in an unknown environment. Firstly, mapping relationships between image features of the servoing object and the joint angles of the robot are derived and learned by a neural network. Secondly, a learning controller based on the neural network is designed for the robot to trace the object. Thirdly, a discrete time impedance control law is obtained for the force servoing of the robot, the on-line learning algorithms for three neural networks are developed to adjust the impedance parameters of the robot in the unknown environment. Lastly, wiping experiments are carried out by using a 6 DOF industrial robot with a CCD camera and a force/torque sensor in its end effector, and the experimental results confirm the effecti veness of the approach.展开更多
Web-log contains a lot of information related with user activities on the Internet. How to mine user browsing interest patterns effectively is an important and challengeable research topic. On the analysis of the pres...Web-log contains a lot of information related with user activities on the Internet. How to mine user browsing interest patterns effectively is an important and challengeable research topic. On the analysis of the present algorithm’s advantages and disadvantages we propose a new concept: support-interest. Its key insight is that visitor will backtrack if they do not find the information where they expect. And the point from where they backtrack is the expected location for the page. We present User Access Matrix and the corresponding algorithm for discovering such expected locations that can handle page caching by the browser. Since the URL-URL matrix is a sparse matrix which can be represented by List of 3-tuples, we can mine user preferred sub-paths from the computation of this matrix. Accordingly, all the sub-paths are merged, and user preferred paths are formed. Experiments showed that it was accurate and scalable. It’s suitable for website based application, such as to optimize website’s topological structure or to design personalized services. Key words Web Mining - user preferred path - Web-log - support-interest - personalized services CLC number TP 391 Foundation item: Supported by the National High Technology Development (863 program of China) (2001AA113182)Biography: ZHOU Hong-fang (1976-), female.Ph. D candidate, research direction: data mining and knowledge discovery in databases.展开更多
A novel construction algorithm is presented to generate a conforming Voronoi mesh for any planar straight line graph (PSLG). It is also extended to tesselate multiple-intersected PSLGs. All the algorithms are guarante...A novel construction algorithm is presented to generate a conforming Voronoi mesh for any planar straight line graph (PSLG). It is also extended to tesselate multiple-intersected PSLGs. All the algorithms are guaranteed to converge. Examples are given to illustrate its efficiency.展开更多
Location based services is promising due to its novel working style and contents. A software platform is proposed to provide application programs of typical location based services and support new applications develop...Location based services is promising due to its novel working style and contents. A software platform is proposed to provide application programs of typical location based services and support new applications developing efficiently. The analysis shows that this scheme is easy implemented, low cost and adapt to all kinds of mobile nework system.展开更多
A new method as a post-processing step is presented to improve the shape quality of triangular meshes, which uses a topological clean up procedure and discrete smoothing interpolate (DSI) algorithm together. T...A new method as a post-processing step is presented to improve the shape quality of triangular meshes, which uses a topological clean up procedure and discrete smoothing interpolate (DSI) algorithm together. This method can improve the angle distribution of mesh element. while keeping the resulting meshes conform to the predefined constraints which are inputted as a PSLG.展开更多
Pulse coupled neural network (PCNN) has a specific feature that the fire of one neuron can capture its adjacent neurons to fire due to their spatial proximity and intensity similarity. In this paper, it is indicated t...Pulse coupled neural network (PCNN) has a specific feature that the fire of one neuron can capture its adjacent neurons to fire due to their spatial proximity and intensity similarity. In this paper, it is indicated that this feature itself is a very good mechanism for image filtering when the image is damaged with pep and salt (PAS) type noise. An adaptive filtering method, in which the noisy pixels are first located and then filtered based on the output of the PCNN, is presented. The threshold function of a neuron in the PCNN is designed when it is used for filtering random PAS and extreme PAS noise contaminated image respectively. The filtered image has no distortion for noisy pixels and only less mistiness for non-noisy pixels, compared with the conventional window-based filtering method. Excellent experimental results show great effectiveness and efficiency of the proposed method, especially for heavy-noise contaminated images.展开更多
In this paper, a novel biometric identification system is presented toidentify a person''s identity by his/her palmprint. In contrast to existing palmprint systems forcriminal applications, the proposed system...In this paper, a novel biometric identification system is presented toidentify a person''s identity by his/her palmprint. In contrast to existing palmprint systems forcriminal applications, the proposed system targets at the civil applications, which requireidentifying a person in a large database with high accuracy in real-time. The system is constitutedby four major components: User Interface Module, Acquisition Module, Recognition Module and ExternalModule. More than 7,000 palmprint images have been collected to test the performance of the system.The system can identify 400 palms with a low false acceptance rate, 0.02%, and a high genuineacceptance rate, 98.83%. For verification, the system can operate at a false acceptance rate, 0.017%and a false rejection rate, 0.86%. The execution time for the whole process including imagecollection, preprocessing, feature extraction and matching is less than 1 second.展开更多
To slove the problem that the implementation of Transmission Control Protocol (TCP) Vegas does not consider the impact of TCP's segment size on Round-Trip Time (RTT) while calculating RTT and the smallest RTT (bas...To slove the problem that the implementation of Transmission Control Protocol (TCP) Vegas does not consider the impact of TCP's segment size on Round-Trip Time (RTT) while calculating RTT and the smallest RTT (base _rtt), this paper proposes a modified congestion control mechanism for TCP Vegas. It first groups segments on the basis of segments size and updates base _rtt based on groups. In congestion avoidance phase, TCP Vegas changes its window size according to the measured RTT of the segment corresponding to the latest received ACK and the base _rtt retrieved from the group that the segment belongs to. Simulations illustrate that the enhanced TCP Vegas can achieve higher throughput up to 40 percent.展开更多
基金This project was supported by the research foundation of China Education Ministry for the scholars from abroad (2002247).
文摘A learning-based control approach is presented for force servoing of a robot with vision in an unknown environment. Firstly, mapping relationships between image features of the servoing object and the joint angles of the robot are derived and learned by a neural network. Secondly, a learning controller based on the neural network is designed for the robot to trace the object. Thirdly, a discrete time impedance control law is obtained for the force servoing of the robot, the on-line learning algorithms for three neural networks are developed to adjust the impedance parameters of the robot in the unknown environment. Lastly, wiping experiments are carried out by using a 6 DOF industrial robot with a CCD camera and a force/torque sensor in its end effector, and the experimental results confirm the effecti veness of the approach.
文摘Web-log contains a lot of information related with user activities on the Internet. How to mine user browsing interest patterns effectively is an important and challengeable research topic. On the analysis of the present algorithm’s advantages and disadvantages we propose a new concept: support-interest. Its key insight is that visitor will backtrack if they do not find the information where they expect. And the point from where they backtrack is the expected location for the page. We present User Access Matrix and the corresponding algorithm for discovering such expected locations that can handle page caching by the browser. Since the URL-URL matrix is a sparse matrix which can be represented by List of 3-tuples, we can mine user preferred sub-paths from the computation of this matrix. Accordingly, all the sub-paths are merged, and user preferred paths are formed. Experiments showed that it was accurate and scalable. It’s suitable for website based application, such as to optimize website’s topological structure or to design personalized services. Key words Web Mining - user preferred path - Web-log - support-interest - personalized services CLC number TP 391 Foundation item: Supported by the National High Technology Development (863 program of China) (2001AA113182)Biography: ZHOU Hong-fang (1976-), female.Ph. D candidate, research direction: data mining and knowledge discovery in databases.
基金Supported by the Science Technology Development Program of Beijing Municipal Education Commission (KM200510011004)
文摘A novel construction algorithm is presented to generate a conforming Voronoi mesh for any planar straight line graph (PSLG). It is also extended to tesselate multiple-intersected PSLGs. All the algorithms are guaranteed to converge. Examples are given to illustrate its efficiency.
基金This work is supported by Korea ITRC developing plan and Sino- Korea GIS center
文摘Location based services is promising due to its novel working style and contents. A software platform is proposed to provide application programs of typical location based services and support new applications developing efficiently. The analysis shows that this scheme is easy implemented, low cost and adapt to all kinds of mobile nework system.
文摘A new method as a post-processing step is presented to improve the shape quality of triangular meshes, which uses a topological clean up procedure and discrete smoothing interpolate (DSI) algorithm together. This method can improve the angle distribution of mesh element. while keeping the resulting meshes conform to the predefined constraints which are inputted as a PSLG.
基金supported by the National Natural Science Foundation of China(Grant Nos.60371044,60071026)the National Visiting Scholar Fund
文摘Pulse coupled neural network (PCNN) has a specific feature that the fire of one neuron can capture its adjacent neurons to fire due to their spatial proximity and intensity similarity. In this paper, it is indicated that this feature itself is a very good mechanism for image filtering when the image is damaged with pep and salt (PAS) type noise. An adaptive filtering method, in which the noisy pixels are first located and then filtered based on the output of the PCNN, is presented. The threshold function of a neuron in the PCNN is designed when it is used for filtering random PAS and extreme PAS noise contaminated image respectively. The filtered image has no distortion for noisy pixels and only less mistiness for non-noisy pixels, compared with the conventional window-based filtering method. Excellent experimental results show great effectiveness and efficiency of the proposed method, especially for heavy-noise contaminated images.
文摘In this paper, a novel biometric identification system is presented toidentify a person''s identity by his/her palmprint. In contrast to existing palmprint systems forcriminal applications, the proposed system targets at the civil applications, which requireidentifying a person in a large database with high accuracy in real-time. The system is constitutedby four major components: User Interface Module, Acquisition Module, Recognition Module and ExternalModule. More than 7,000 palmprint images have been collected to test the performance of the system.The system can identify 400 palms with a low false acceptance rate, 0.02%, and a high genuineacceptance rate, 98.83%. For verification, the system can operate at a false acceptance rate, 0.017%and a false rejection rate, 0.86%. The execution time for the whole process including imagecollection, preprocessing, feature extraction and matching is less than 1 second.
文摘To slove the problem that the implementation of Transmission Control Protocol (TCP) Vegas does not consider the impact of TCP's segment size on Round-Trip Time (RTT) while calculating RTT and the smallest RTT (base _rtt), this paper proposes a modified congestion control mechanism for TCP Vegas. It first groups segments on the basis of segments size and updates base _rtt based on groups. In congestion avoidance phase, TCP Vegas changes its window size according to the measured RTT of the segment corresponding to the latest received ACK and the base _rtt retrieved from the group that the segment belongs to. Simulations illustrate that the enhanced TCP Vegas can achieve higher throughput up to 40 percent.
