Applied research presented in this paper describes an approach to provide meaningful evaluation of the Machine Learning (ML) components in a Full Motion Video (FMV) Machine Learning Enabled System (MLES). The MLES itself is not discussed in the paper. We focus on the experimental activity that has been designed to provide confidence that the MLES, when fielded under dynamic and uncertain conditions, performance will not be undermined by a lack of ML robustness. For example, to real-world changes of the same scene under differing light conditions. The paper details the technical approach and how it is applied to data, across the overall experimental pipeline, consisting of a perturbation engine, test pipeline and metric production. Data is from a small imagery dataset and the results are shown and discussed as part of a proof of concept study.
Traditionally, the production of high quality Synthetic Aperture Radar imagery has been an area where a potential user would have to expend large amounts of money in either the bespoke development of a processing chain dedicated to his requirements or in the purchase of a dedicated hardware platform adapted using accelerator boards and enhanced memory management. Whichever option the user adopted there were limitations based on the desire for a realistic throughput in data load and time. The user had a choice, made early in the purchase, for either a system that adopted innovative algorithmic manipulation, to limit the processing time of the purchase of expensive hardware. The former limits the quality of the product, while the latter excludes the user from any visibility into the processing chain. Clearly there was a need for a SAR processing architecture that gave the user a choice into the methodology to be adopted for a particular processing sequence, allowing him to decide on either a quick (lower quality) product or a detailed slower (high quality) product, without having to change the algorithmic base of his processor or the hardware platform. The European Commission, through the Advanced Techniques unit of the Joint Research Centre (JRC) Institute for Remote Sensing at Ispra in Italy, realizing the limitations on current processing abilities, initiated its own program to build airborne SAR and Electro-Optical (EO) sensor systems. This program is called the European Airborne Remote Sensing Capabilities (EARSEC) program. This paper describes the processing system developed for the airborne SAR sensor system. The paper considers the requirements for the system and the design of the EARSEC Airborne SAR Processing System. It highlights the development of an open SAR processing architecture where users have full access to intermediate products that arise from each of the major processing stages. It also describes the main processing stages in the overall architecture and illustrates the results of each of the key stages in the processor.
The aim of the research described in this paper has been to develop change detection algorithms which would be suitable for an operational map update system. Two algorithms have been developed, one based on edge matching and another on object analysis. These algorithms have been tested on thematic mapper imagery, using ordnance survey (OS) and forestry commission (FC) digitized maps. When possible, knowledge of forest practice has been used to augment the results. Results indicate that, with minimal user interaction, major areas of change can be determined. Increasing user intervention can reduce algorithm complexity and processing time, at the expense of automation.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.