Characterization and acceleration of a global motion mapping system

InSAR is a technique used by radar-based remote sensing for applications such as tracking spatial displacement/deformation of an object on earth, e.g., building and a bridge, over time. The object analysis part of InSAR processing belongs to an offline processing chain that operates on very high resolution images consisting of huge volume of data. The processing chain consists of several independent or loosely coupled operating entities with different resources requirements. Often multiple products are required to enter the chain simultaneously and they share processing resources. Manually scheduling and managing atomic tasks on shared resources are both complex and susceptible to errors. The primary goal of this project is to characterize kernels involved in each step in the processing chain in terms of resources usage, data volumes, and scalability, and requirements analysis and technology assessment for provisioning a task and resources scheduling middleware. This project would benefit 3vG in terms of improving efficiency of their processing chain, understanding scalability requirements and help making future hardware decisions.

Faculty Supervisor:

Dr. Matei Ripeanu

Student:

Tahsin Arafat Reza

Partner:

3v Geomatics

Discipline:

Engineering - computer / electrical

Sector:

Information and communications technologies

University:

University of British Columbia