Maxar Student Lightning Talk Session
February 23, 2024
9:30 - 10:15 amLocation: Center Ballroom
Three students from participating Southern California GIS academic programs present five-minute lightning talks, followed with discussion and Q&A with GIS experts and Summit participants
Industry Panel:
- Barry Tilton, Technology Evangelist, Maxar
- Cora Chong, GIS Specialist, Innovation, CBRE
- Christine Devine, Director of Education and Professional Development, United States Geospatial Intelligence Foundation
Student Presenters:
John Adrian
California State University, Long Beach
“Preprocessing Imagery for Mapping the Martian Landscape using GPT-4 and Python Scripting”
The Martian geospatial landscape presents unique challenges in planetary mapping and analysis. This project introduces an innovative alternative workflow for the automatic coregistration of high-resolution Martian imagery based on the methodology of Sidiropoulos et al. (2018). This effort, in collaboration with GPT-4 as a Python coding expert and software development guide, focuses on automating the coregistration of CTX and THEMIS imagery. Aimed at scaling the process to consumer-level computing capabilities, it broadens accessibility to automatic image co-registration tools for GI-Science. Integrating the Scale-Invariant Feature Transform (SIFT) for feature detection and employing Facebook AI Similarity Search (FAISS) for efficient feature matching, we diverge from the original coupled decomposition method for scalability. Additionally, we are exploring the implementation of the Ring Matching algorithm as an intermediary step to refine SIFT keypoints selection for the FAISS matching process. Following matching, the Random Sample Consensus (RANSAC) algorithm calculates a new transformation model. This model is then applied to the georeference data in ArcGIS Pro, finalizing the alignment of high-resolution Martian images.
This approach not only enhances the efficiency and accuracy of geological mapping but also democratizes high-level Martian geospatial analysis, making advanced techniques more feasible for researchers with limited computing resources.
Leo Lerner
Ph.D. candidate, Population, Health and Place, USC Spatial Sciences Institute
“Affective geographic information science: can visual carto-elicitation evoke the sensorimotor?”
By using innovative ethnographic methods, the project examines the distribution of power and knowledge in social science research outcomes. Vision and visuality are powerful mediators of culture and ideology, but visual methods and qualitative GIS have not been used together in a multimodal project, despite the abundance of geotagged images created and shared daily on mobile phones. Mixing methods like this as part of a new public participatory geographic information science (PPGIS) presents the potential to triangulate new knowledges which are literally unforeseen. Materials are produced by the researcher and the participant in collaboration, and then reproduced for discussion in a semi-structured interview. By discussing the spatial data as interrelated with the content and meaning of the images, new forms of representation and meaning emerge.
This is a multimodal approach which despite great potential, has been only superficially explored. One of the central questions asks to what extent visual materials can be incorporated into an established, conventional, platform like a GIS. Investigating this tension, and using it to initiate new interpretations and conclusions, should suggest areas for further analysis, or recast the framing of qualitative research questions altogether.
Priyanka Somai
University of Redlands
“Digital Inclusion in India: A Geospatial and Multivariate Perspective”
The lightning talk addresses the digital divide within India, a challenge marked by the unequal distribution of digital resources and internet access. This pervasive issue has far-reaching consequences, leading to disparities in education, healthcare, and economic opportunities.
Our objective is to analyze India’s dynamic technology landscape, including personal computer and mobile phone ownership and access, internet and broadband subscriptions, and their purposeful uses. Geostatistical and regression analysis reveal the geographical patterns and disparities influencing technology adoption and use and its relationship with demographics, economics, social dynamics, and infrastructure factors.
Using GIS, we map internet connectivity, purposeful use, and digital infrastructure across India’s 29 states. These maps help to identify regions with limited technology access and provide visual cues about clustering and agglomeration of technology access and use. Clusters can be further characterized by their attributes and interpreted by referring to prior research studies.
We utilize geostatistical tools and multivariate methods to correlate technology access with socioeconomic factors such as gender, age, income, education, infrastructure, innovation, and social capital.
By identifying and analyzing underserved regions, the country can develop policies and initiatives to ensure equitable technology access, enhancing the quality of life for all and fostering inclusive growth across the nation.
Speakers
Barry Tilton is the Technology Evangelist for Strategic Growth at Maxar Technologies, and the Director of the Mid-Atlantic Region of IEEE and IEEE-USA. He has over 35 years’ experience as an Electrical Engineer in the aerospace field, and is both a Licensed Professional Engineer, and Certified Project Management Professional. He is a 20-year veteran of the US Air Force, where he had over 1000 flight hours as a test director, and has worked as a Chief Engineer, and Chief Scientist, and Program Director for the US Defense and State Departments as well as several international efforts with countries on five Continents. In industry he has served as a CTO, Vice President of Engineering, Chief Architect and Division Manager. Barry earned his BSEE from USC and his MSEE from Northrop University. He is a Senior Member of IEEE, Eta Kappa Nu and the AIAA, and a Fellow of the American Geographical Society.
Cora Chong is a GIS Specialist on the Innovation team within the Location Intelligence group at CBRE, a global real estate services company. Her work focuses on spatial consulting for high-value, high-profile commercial real estate projects, which may include adaptive reuse, mixed use, entertainment, and placemaking components. She graduated from the University of Southern California with a Master’s of Science in Geographic Information Science and Technology, and is also a GIS Instructor at Cuesta College.
Christine Devine, the Director of Education and Professional Development at the United States Geospatial Intelligence Foundation (USGIF). She lead’s USGIF academic efforts including the collegiate accreditation program, scholarship program, and K-12 geospatial educational outreach. She is a human geographer by training with a M.S. in Geography from George Washington University.