NCSI: UAV Assisted Pathfinding in Flooding Area

Status	Completed
Mentor Name	Jiang Lu
Mentor's XSEDE Affiliation	Campus Champion
Mentor Has Been in XSEDE Community	1-2 years
Project Title	UAV Assisted Pathfinding in Flooding Area
Summary	This is follow-up research of Unmanned Aerial Vehicle (UAV) for Object Tracking. With the help of HPC, more UAV based applications can be possible. In this project, we will continue using computer vision techniques for optimal pathfinding with the assistance of UAV. UAV can capture wider view than humans. Thus, it provides better solution for a path from source to destination. This optimization problem has many real applications such as helping rescue in flooding, fire or earthquake. The students will be asked to evaluate the pathfinding algorithms in XSEDE HPC and to implement with UAV that has been built with previous funding support.
Job Description	The ongoing research investigates the optimal pathfinding algorithm using image processing with assistance from UAV. The application can be flooding rescue. The students will be responsible for working a full spring semester for 10 hours per week. The students will participate in weekly group meetings for the monitoring of the project. The students will research existing methods of pathfinding algorithms. The students will evaluate the methods and pick one for our UAV assist pathfinding project. The students will implement the picked method in HPC for modeling. The students will have access to the center for robotic software at University of Houston Clear Lake. The students' final project will be used for a further research project in another semester.
Computational Resources	XSEDE training classes XSEDE supercomputing Clusters with Computer Vision AI/Deep learning software Data visualization software Additionally, the mentor can use the resources in the center for robotic software at the University of Houston Clear Lake.
Contribution to Community	The real application can be rescue in flooding, hurricane, fire or earthquake. Especially in golf coast area, such as Houston, UAV can be used for hurricane caused flooding and rescue.
Position Type	Learner
Training Plan	The student will be supervised by the mentor and graduate students, who have experience in image processing. The student will learn the basics of deep learning, computer vision. The student will be trained for the usage of XSEDE at University of Houston Clear Lake and online.
Student Prerequisites/Conditions/Qualifications	1. Interest and background in machine learning and image processing. 2. Thrive on the challenge of complex problem-solving. 3. Effective communication. 4. At least a strong background in one of the below, or a combination two having a solid background from the below list: Programming, Robotics, Mathematics, Deep Learning, Image Processing, and Computer Vision.
Duration	Semester
Start Date	01/19/2021
End Date	05/19/2021