Guannan Lu was a graduate student in ViGIR Lab at University of Missouir-Columbia, pursuing his Master Degree of Science of Electrical Engineering. He explored a number of different fields in Computer Science as well as Electronic Engineering and discovered his strong interest in computer vision, machine learning and robotics. From 2012 to 2014, he worked at ViGIR Lab as Graduate Research Assistant, supervised by Professor Guilherme DeSouza.

Publication

u7_start u7_end u7_line

Contact

u15_start u15_end u15_line

Research Assistant at ViGIR Lab

Guannan Lu

u20_start u20_end u20_line

Project

Publication

Lu, G., DeSouza, G. N., Armer, J., and Shyu, C. "A Self-Monitoring System for Early Detection and Management of Lymphedema Using Smart Phones". ILF, 2014. (accepted)

 

 

Lu, G., DeSouza, G. N., Armer, J., and Shyu, C. "A New Algorithm for 3D Resigtration and its Application in Self-Monitoring and Early Detection of Lymphedema". Journal Innovation and Research in BioMedical engineering, from the AGBM (Alliance for engineering in Biology and Medicine), Elsevier, 2014. (Reviewing)

 

 

Lu, G., DeSouza, G. N., Armer, J., and Shyu, C. "Comparing Limb-Volume Measurement Techniques: 3D Models from an Infrared Depth Sensor versus Perometry and Water Displacement" in 15th IEEE International Conference on e-Health Networking, Application and Services, Lisbon, Portugal, Oct. 9-12, 2013. (Same paper to appear in the Journal of Lymphology)

 

 

Lu, G., DeSouza, G. N., Armer, J., Shyu, C., and Anderson, B., "Comparing Limb-Volume Measurement Techniques: 3D Models from an Infrared Depth Sensor versus Perometry and Water Displacement" in 24th Congress of the International Society of Lymphology, Rome, Italy, Sept. 16-20, 2013

 

 

Lu, G., DeSouza, G. N., Armer, J., Anderson, B., and Shyu, C. "A System for Limb-Volume Measurement using 3D Models from an Infrared Depth Sensor" in IEEE Symposium Series on Computational Intelligence, Symposium on Computational Intelligence for Health Care, CICARE-13, April, 2013

A System for Limb-Volume Measurement

In this project, we present a method for measuring limb volume and for detecting early swelling associated with lymphedema. The system relies on IR imaging sensors, such as in the Microsoft Kinect. This technique will allow for the future development of tools for self-management and specialist monitoring, and when compared to other commercially available devices, our system is less expensive, equally or more reliable/accurate, and much more user friendly.

Project

u37_start u37_end u37_line

Contact

u41_start u41_end u41_line

LGNOFFICIAL[AT]GMAIL.COM

3D Reconstruction based on Smartphone

In this project, a new 3D reconstruction system  based on smart phone is proposed. The system recovers the camera positions to a global unit scale by combining the data from a gyroscop scnsor and a normal RGB smart phone camera. Accuracte 3D models can be reconstructed by low resolution images (640*480) . This system will be implemented as an smart phone application as part of  American Lymphedema Framework Project (ALFP).

Surveillance Unmanned Ground Vehicle (Bachelor Capstone Project)

The goals of the SUGV are to provide the user with information about a given area by patrolling its given boundaries, and to provide to the user a system that independently charges when necessary. In this project, the GPS guidance guides the SUGV along its path and image processing is used to accurately position the SUGV in its optimal position within the garage.

Deep Learning for 3D Object Recognition

This is one part of "Comparison of Supervised Learning Techniques used for 3D Object Recognition". The choice of the right classifier is of paramount importance and might improve or worsen the recognition results. For this specific purpose, we are proposing in this work to study the effect of the classifier chosen on the recognition.

My contribution in this project is the implementation of the deep learning (DP) techenique. DP is achieved by pre-training for each layer. Each layer is trained with RBM sequentially. Then, a fine-tune is applied on all parameters of the previously built model.

Last Modified: Jul-21-2014