John Dolloff will visit the U of A campus on Monday, Nov. 11, 2024, to meet with research faculty and staff at the Center for Advanced Spatial Technologies (CAST) and to present a one-hour lecture on "MIN-based Fusion: Sequential block adjustments integrated with a self-generating ground control network (aka Metric Information Network) - An advanced application of 'rigorous error propagation.'"

During his lecture, Dolloff will address the difficulties associated with maintaining and updating a modern geodetic reference frame based on high-resolution satellite images and present a consolidated and consistent method for not only estimating locations of reference points in these frameworks but updating the uncertainty of the points as new information is added to the network. These geodetic reference frames are the foundation of all map-based activities from personal navigation to all forms of transportation and logistics.

The presentation will include general principles of rigorous error propagation, which involves proper modeling of all applicable errors, appropriate use of predictive statistics as well as sample statistics, transformation between predictive statistics, optional Monte-Carlo simulation of errors, estimators such as Weighted Least Squares and Kalman Filters and their impacts on geospatial applications involving expression of reliable predicted accuracy and optimal geolocation extractions such as block adjustments.

Dolloff will then present the MIN-based fusion which sequentially incorporates information from various image block adjustments into a self-generating and expanding network of ground control points that generally increases in area of coverage and that also increases in accuracy. The network also includes a rigorous full error covariance matrix for all points, including the cross-covariances (correlations) between all point pairs.

Presentation details:

• Date: Monday, Nov. 11
• Location: JBHT 535 (5th Floor Boardroom)
• Time: 2-3 p.m.

Note: This is an in-person event only.

Presenter Bio: Dolloff has more than 40 years of experience in geospatial science and remote sensing, specializing in photogrammetry and applications of advanced linear algebra, estimation theory, and probability and statistics. He is currently employed at KBR and was previously employed at BAE Systems and its legacy companies where he was a technical director and an engineering fellow. He holds a bachelor's degree in mathematics and a master's degree in applied mathematics.

Dolloff is credited with the invention of the Replacement Sensor Model (RSM) and the Metric Information Network (MIN). RSM is operational today and provides full functionality (ground-to-image relationship, adjustability, predicted accuracy) equivalent to an arbitrary but original (physical) sensor model from which it is generated. As such, a "down-stream" user only requires the use of one sensor model (RSM) for exploitation of those imaging systems which provide corresponding RSM support data.

---

About CAST: CAST acquires, aggregates, analyzes, and applies geospatial data. Whether reconstructing past and present environments or reconciling virtual and physical worlds, CAST approaches spatial data problems in a project-based, applications-oriented, collaborative manner. CAST has resided in the J. William Fulbright College of Arts and Sciences since its founding in 1991 and is actively involved in extensive research efforts. CAST is principally supported by competitive grants from federal and state agencies that are often developed in collaboration with industry and academic partners. CAST offers students, faculty, and the public opportunities to learn about and use a wide array of geospatial technologies including sUAS airframes and sensor platforms, terrestrial and airborne LiDAR, GNSS, photogrammetry, and advanced processing software.