Qifeng (Luke) Lu, PhD

Researcher

Phone: +1 (703) 518-2972 [office]


Dr. Lu gained his PhD from the multidesciplinary Geospatial Engineering program at Viriginia Tech, and Master's degrees in Computer Science and Transportation Engineering at University of Massachusetts Amherst.

Dr. Lu is specialized in best first searches. His dissertation is on bivariate best first searches to process multi-category based queries in a graph and their applications in GIS-T. His work significantly extends the functionality of best first searches. He proposed multivariate best first searches to process queries with multiple categories/points of interest, and created a set of novel challenges both in research and in practice.

Dr. Lu joined MacroSys as a Research Analyst/GIS Specialist in July, 2009. He continues his work in best first search and takes a strong role in implmenting GISs that use best first searches for optimization and to provide advanced routing services in transportation, logistics, and computer networking.


Research Interests:

Dr. Lu has broad interests in GIS-T, informed search, optimization, logistics, Location Based Services, data mining, wireless and geosensor network, and transportation engineering.


Publications:

    Journal papers
  1. Qifeng Lu,Yao Liang. Multiresolution Learning on Neural Network Classifiers: A Systematic Approach. Submitting to the journal NeuralComputing (0.97).
  2. Qifeng Lu, Feng Chen, Kathleen Hancock. On Path Anomaly Detection in a Large Transportation Network. Submitted to the Journal Computers, Environment, and Urban Systems (under second-round review, impact factor: not available).
  3. Taihua Yang, Qifeng Lu, Enzhi Wang. Analysis of Three-Dimensional Seepage Of Bhote Koshi Hydroelectric Project In Nepal, Journal of Engineering Geology, 4(04), 2000.
  4. Enzhi Wang, Qifeng Lu, Xiong Sun. Impervious Wall Optimization Design of Gravity Dam on Pebble Foundation, Chinese journal of Water Resources and Hydropower Engineering, 30(12),1999.
  5. Enzhi Wang, Qifeng Lu, Wenwei Li. Expert System for Concrete Quality Control, Chinese journal of Water Resources and Hydropower Engineering, 30(5), 1999.
  6. Peer reviewed published/accepted conference papers
  7. Qifeng Lu, Yao Liang. Multiresolution Learning on Neural Network Classifiers: A Systematic Approach. To apprear: Proceedings of 12th International Conference on Network-Based Information Systems. Indianapolis, IN, USA,August 19-21, 2009
  8. Qifeng Lu, Stephen Sedlock. Development of Trail Network Model and a Web-based Bikeway Routing Service System, Proceedings of 16th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM GIS 2008), Page 87-88. Irvine, CA, Nov., 2008.
  9. Qifeng Lu, Kathleen Hancock, GMLinAir: Enabling Efficient and Power-conserving Wireless GML Communication and Visualization, 2nd International Conference on Geosensor Networks (GSN'06), Boston, USA, 2006.
  10. Qifeng Lu, Kathleen Hancock, TagCMP: Enabling Efficient GML Communication with Power Conservation in a Wireless Environment, 2nd International Conference on Geosensor Networks (GSN'06), Boston, USA, 2006.
  11. Erik Learned-Miller, Qifeng Lu, Angela Paisley, Peter Trainer, Volker Blanz, Katrin Dedden, and Ralph E. Miller. Detecting Acromegaly: Screening for Disease with a Morphable Model. Medical Image Computing and Computer-Assisted Intervention (MICCAI), Volume 2, pp. 495-503, 2006.
  12. Qifeng Lu, Kathleen Hancock. Micro Simulation of Large Truck Activities at Tight Intersections.TRB Compendium of Papers. 85th Annual Meeting on January 22-26, 2006.
  13. Purushottam Kulkarni, Deepak Ganesan, Prashant Shenoy and Qifeng Lu. SensEye: A Multi-tier Camera Sensor Network, Proceedings of the 13th annual ACM international conference on Multimedia, P229-238, Sinapore, 2005.(Best Student Paper Finalist).
  14. Erik Learned-Miller, Qifeng Lu, Angela Paisley, Peter Trainer, Volker Blanz, Katrin Dedden, and Ralph Miller. (2005) "Early Diagnosis of Acromegaly by Facial Pattern Recognition." Abstract for the Ninth International Pituitary Congress, San Diego, CA, USA, June 7-9, 2005.
  15. Qifeng Lu, Enzhi Wang. Impervious wall optimization design of gravity dam on pebble foundation. Proceedings of International Conference on Anchoring & Grouting towards the new Century, Guangzhou: Zhongshan University Publisher, Oct., 1999, p292-295.
  16. Other publications
  17. Qifeng Lu, Kathleen Hancock. Data Oriented Approximately Trip Planning Query Processing on Consumer Destinations within a Dense Road Network (abstract), Association of American Geographers (AAG) Annual Meeting, Boston, USA, 2008.
  18. Qifeng Lu. Screening Patients for Acromegaly by Examining Facial Photographs. M.S. Thesis in Computer Science Department, University of Massachusetts Amherst, 2005.
  19. Qifeng Lu, Erik Learned-Miller, Angela Paisley, Peter Trainer, Volker Blanz, Katrin Dedden, and Ralph Miller. (2005) "Detecting Acromegaly: Screening for Diseases with a Morphabel Model." UMass Amherst Technical Report 05-37, 2005.
  20. Joseph DeStefano, Qifeng Lu and Erik Learned-Miller, (2005) "A Probabilistic Upper Bound on Differential Entropy" UMass Amherst Technical Report 05-12, 2005.
  21. Qifeng Lu. Micro Traffic Simulation of Truck Activities at Tight Intersections on Surface Streets. M.S. Thesis, University of Massachusetts Amherst, 2004.
  22. Patents

  23. Qifeng Lu, Kathleen Hancock. O*: A Bivariate Best First Search Framework, and O*-SCDMST to Process Optimal Sequence Traversal Query in a Graph for Trip Planning. VTIP:09-126.
  24. Qifeng Lu, Kathleen Hancock. C*: A Bivariate Best First Search Framework, and C*-P and C*-Dijkstra, to Process Category Sequence Traversal Query in a Graph for Trip Planning. VTIP:09-125.

Selected Projects:

    National Corridor Freight Performance Measure Web System

    The web system is developed to provide FHWA, State DOT, MPOs, and researchers freight performance measures along national corridors. Queris over time and locations can be customized and linear referencing is used to provide performances over 3-mile segments with ArcGIS server.

    Real Time Micro Traffic Simulation

    It is crucial to incorporate real time traffic information into micro simulations to timely manage event-related activities in traffic operations. This project used TransModeler and real time data sets to monitor possible events along the corridor Route 50 at Arlington, VA.

    Micro Traffic Simulation of Truck Activities

    Transportation professionals have long realized the importance of incorporating truck characteristics into transportation design and traffic operations. Large trucks differ dramatically from passenger cars in many physical and operational characteristics. They are longer, wider, heavier, less maneuverable, and have slower acceleration and deceleration rates and higher emissions outputs than passenger cars. Thus they can cause traffic disturbances leading to the generation of shock waves, traffic delays, adverse variations in vehicle speeds, etc. In addition, their larger size and inertia pose safety concerns in case of crashes and may influence the response of passenger vehicles drivers. On the other hand, their movements also can be impacted by road conditions such as intersection geometric characteristics. These impacts can be more serious and complex at tight intersections on surface street networks. This project prevides a framework to simulate the impacts of truck activities at tight intersections on surface streets, proposes simulation methods to quantify those impacts, and presents the results from the study on those impacts on the south Boston local street network near the Boston Container Terminal using a modified surface street network simulation model in Traffic Software Integrated System (TSIS) 5.1.

    Developing a Logical Model for a Geo-Spatial Right-of-Way Land Management System

    Right-of-way (ROW) requirements are significant components of project schedule and cost. Manually recorded ROW information includes agency ownership, appraisal information acquisition status, and property -management functions that are important for addressing real estate issues, utilities, environmental permitting and mitigation, access management, outdoor advertising control, and programming. Electronic management of this information improves coordination and consistency of data, leading to reduced project delivery delays caused by ROW acquisition. In addition, the ability to retrieve these data electronically provides fast, convenient, and consistent access to all users, reducing the time and expense needed to ship documents; eliminating repetitive entries; minimizing data-entry errors caused by multiple formats; and ultimately saving money for the DOTs. Electronic management of real estate information could improve coordination with local jurisdictions and provide appropriate data to the public on agency ownership of property.


    The automation of ROW functions and development of data-integration models using existing technology, including geo-spatial applications, are needed to enable multiple users to access ROW information quickly and easily.


    Biker Routing on a Trail Network

    To facilitate the decision making of a state Department of Transportation (DOT) on non-motorized transportation assets, a transportation network model, TNM, is developed. Based on this model, a web-based routing service is implemented to provide bikers routing services on the trail network.

     

    Alexandria Archeology GIS Mapping

    Digitally preserve/rectify important historic maps for the Alexandria Archaeology Museum beginning with the area outside of Old Town Alexandria where archaeological resources are most threatened by expanding development.Deploy the digital maps through arcIMS.

    SenseEye

    SenseEye is a multi-tier network of heterogeneous wireless nodes and cameras, which was employed to show a multi-tier sensor network can reconcile the traditionally conflicting systems goals of latency and energy efficiency. A surveillance application was employed using SensEye comprising three tasks: object detection, recognition and tracking. Novel mechanisms were designed for low-power low-latency detection, low-latency wakeups, efficient recognition and tracking. An experimental evaluation shows that, when compared to a single-tier prototype, our multi-tier SensEye can achieve an order of magnitude reduction in energy usage while providing comparable surveillance accuracy.

    Acromegaly

    Acromegaly is a hormonal disorder. Acromegaly means extremities and enlargements in Greek. Its typical symtoms are disfiguring growth of the bones of the skull and swelling of the face, hands, and feet. A facial features enlargement is the second most prevalent symptom with a prevalence over 95%. It includes protruded jaws, eye brows, and cheekbones, and enlarged lips and nose. The goal in this project is to detect acromegaly automatically from generic frontal facial photographs so that it can be diagnosed and recognized in a timely manner. In collaboration with Volker Blanz, Erik G. Learned-Miller and others, a classification system which prescreens patients for acromegaly using the Morphable model and Support Vector Machines (SVMs) was developed.

    Be aware: The following image can only be used for research purpose! Violation may result in serious consequences including but not limited to lawsuits!

     

    A Probabilistic Upper Bound for Differential Entropy

    To provide a probabilistic MAXIMUM for differential entropy.


Below are some links to some friends, co-authors, and affiliated groups.



 

Last update: 05/12/2009