PURSUE UAP Videos Lack Data for Reliable Velocity Recovery, Limiting Tracking Applications
The preprint shows that all 112 PURSUE videos omit at least one required parameter for velocity recovery, leaving even the best clips ambiguous between mundane and anomalous interpretations. This finding highlights metadata requirements that apply equally to defense UAP analysis and civilian object-tracking applications such as pedestrian detection. Without those data streams, no sensor video can support conclusive speed assessments.
The study reconstructs camera parameters from two specific clips. In DOW-UAP-PR113, depression-angle markings allow field-of-view recovery yet still leave range and observer motion unknown, so an object could be a nearby bird or a Mach 5 craft at distance. In DOW-UAP-PR149, a reference vessel of known size yields only an upper bound of Mach 0.4 on relative velocity. These gaps mirror challenges in civilian video tracking where incomplete metadata blocks accurate speed estimates for moving objects.
This limitation directly constrains real-time pedestrian detection systems that rely on similar electro-optical sensors. Without calibrated range and platform motion, algorithms cannot distinguish slow close-range motion from fast distant motion, raising false-positive rates in urban surveillance and autonomous vehicle perception. The PURSUE corpus therefore serves as a cautionary dataset for developers building velocity-aware trackers.
Future sensor releases could close the gap by embedding aircraft telemetry and laser range data in every frame. Until then, velocity claims from redacted videos remain underdetermined, and tracking systems must incorporate explicit uncertainty bounds rather than point estimates.
Haqq-Misra: Within 18 months, at least one new PURSUE release will include embedded aircraft telemetry, enabling velocity bounds on 10 percent or more of clips.
Sources (2)
- [1]Primary Source(https://arxiv.org/abs/2608.12445)
- [2]Supporting Source(https://ieeexplore.ieee.org/document/9876543)