High Performance Image Processing Algorithms for Current and Future Mastcam Imagers

Project: Research project

Description

Dr. Jim Bell and Hannah Kerner, a Ph.D.-level graduate student in the School of Earth and Space Exploration at Arizona State University, will conduct research toward several objectives described in this proposal. Dr. Bell will oversee the graduate student and ensure that all deadlines and deliverables are produced and are provided to Applied Research. The work will be conducted over the course of the first 18 months of the project (project start assumed to be 09/01/2017).
The first 6-7 months of the project will be spent developing algorithms and code towards the data clustering and anomaly detection objectives. ASU will explore approaches primarily based on deep neural network and spectral clustering methods to complement the methods explored by Applied Research. Following algorithm development, ASU will perform extensive user studies with scientists on the Mastcam and Mastcam-Z science teams to determine the performance and effectiveness of the algorithms in practice. This testing and algorithm revision cycle is anticipated to take 4 months and be completed by 08/2018. The next phase of ASUs work, in close collaboration with Applied Research, will focus on implementing the algorithms developed for data clustering and anomaly detection as standalone IDL modules. This work will be performed over 4 months and be completed by 12/2018. Having completed the anomaly detection and data clustering research for this study, ASU will work with Applied Research on the evaluation of compression algorithms on Mastcam and Mastcam-Z images using intelligent systems developed at ASU to characterize the perceived effect of compression on scientific images. This work will be performed in the remaining 3 months of the 18 months ASU will contribute to this project (completed by 3/2019).
ASU will contribute details of the proposed work including methods and results to all midterm, annual, and final reports required for the project. In order to disseminate the results of this research to the scientific community, ASU will write conference abstracts and journal papers to present the proposed work at relevant planetary science and computer science conferences and peer-reviewed journals.
StatusFinished
Effective start/end date10/1/179/30/19

Funding

  • National Aeronautics Space Administration (NASA): $85,355.00

Fingerprint

Image sensors
Image processing
Students
Intelligent systems
Computer science
Earth (planet)
Testing