The Need for Standards in UAV Imagery Analysis

doi:10.1117/3.2239237.ap4

Book: Getting Started with UAV Imaging Systems: A Radiometric Guide

Author(s): Barbara G. Grant

Published: 2016

https://doi.org/10.1117/3.2239237.ap4

Author Affiliations +

Abstract

Consider the following fact pattern: an argument supporting the use of armed drones to enforce domestic law, a focus on city monitoring and crowd surveillance by UAV instrument vendors, and increasing sophistication in studies of crowd behavior.5 One logical conclusion deduced from this pattern is that armed drones will be used in domestic law enforcement; it is only a question of when and under what circumstances their use will occur. While a debate on civil liberties is best conducted in a different venue, the essential point is that imagery from UAV sensors will be used to prove or disprove specific assertions surrounding the use of force. This is not as straightforward as it may sound, as two imagery analysts looking at the same image data do not always reach the same conclusions; moreover, no standards exist to judge analysis quality. Even the NIIRS and V-NIIRS ratings discussed in this text require subjective evaluation, albeit by trained analysts. Although establishing an analytical connection between subjective ratings and instrument and environmental parameters6 helps reduce some of the uncertainty regarding data quality, there are many improvements that can be made to enhance the quality of analysis.