The following content was derived from the Foundations of Geographic Information and Spatial Analysis Boot Camp(No link is currently available) by Penn State's College of Earth and Mineral Sciences and is licensed under CC BY 3.0.
As I said on the previous page, both military and civilian missions use Drones or UAVs for reconnaissance and surveillance. These systems often deliver real-time video. Real-time video capabilities, referred to as "motion imagery" or "full motion video" (FMV) are expanding the role of remote sensing and GEOINT in high-tempo military operations and civilian applications. The conventional sensors and platforms presented in previous sections provide critically important, spatially accurate, base map layers for geographic information systems and traditional image analysis. FMV sensors and platforms open the door for persistent surveillance of pinpointed targets on the ground, tracking them as they move and fusing intelligence from other sources to support immediate action. FMV presents significant new challenges to geospatial infrastructure—hardware, software, and analysts.
According to the Motion Imagery Standards Board (MISB) a motion imagery system is "any imaging system that collects at a rate of one frame per second (1 Hz) or faster, over a common field of regard." While MISB makes no formal distinction between motion imagery and full motion video, FMV is generally regarded as "that subset of motion imagery at television-like frame rates (24 - 60 Hz)."
The key phrase "persistent surveillance" provides the important distinction between FMV and traditional analysis of discontinuous imagery. It connotes "constant stare," the ability to watch a point on the ground or to follow a moving target for a long period of time, without interruption. Contrast this need with the typical field of view and revisit time for traditional airborne or spaceborne remote sensing systems and you will begin to appreciate the paradigm shift in geospatial intelligence being stimulated by FMV technology. Traditional imagery analysis tends to be feature-based, focusing structural features of buildings or identification of known objects of interest, such as tank formations and fleets of ships or aircraft. FMV, on the other hand, is activity-based, focusing on capturing the movements of individual people and vehicles or the "patterns of life" observed by small groups (Copeland, 2009).
FMV technologies provide the capability to monitor high-interest activities, including "tracking moving, fleeting, and emerging targets as well as observation of rapidly developing events." (ASPRS, 2009) The phrase "find, fix, and finish" neatly describes the tactical advantage imparted to those who possess this powerful intelligence tool.
Terms commonly used when performing or discussing persistent surveillance are defined by Copeland (2009) as:
- Wide Area
- A region large enough to capture targeted activities of networks
- No specific scale or dimensions, rather region of interest is determined by mission requirements
- Actors and Entities
- An actor is a specific individual
- An entity is an non-human item of interest, such as a car, building, or location
- Activities and Transactions
- An activity is any specific behavior, where the nature of that activity is used to identify actors as enemies.
- A transaction is an observed relationship between entities and actors which allows the designation of enemy
- Consists of actors and/or entities connected by a series of transactions
- A relationship between multiple actors or entities that has been identified by a series of transactions
Digital video cameras, infrared, multispectral, and hyperspectral systems can all be adapted for an FMV application. Because these systems are primarily intended for human interpretation in surveillance applications, where limiting the size of the dataset is needed to facilitate real-time streaming and processing, they tend to have lower spatial resolution and smaller fields of view than their traditional remote sensing counterparts. As with conventional systems, the instantaneous field of view and scale of the resulting imagery will be determined by the operational altitude above ground level (AGL) and the focal length of the camera lens.
The most common sensors comprising an FMV system are a electro-optical (EO) panchromatic digital video camera and an infrared video camera. The EO sensor is simply either a panchromatic or color digital video camera using daylight as the illumination source and recording data in the visible (red to blue) part of the electromagnetic spectrum.
The infrared video (IR) camera acquires thermal imagery, which is useful for detecting thermally emissive objects (e.g. people, running vehicles, etc.) or for collecting imagery at night when there is no ambient light available for the EO camera. The IR cameras can be set to collect "white hot," where the hottest objects in the scene are depicted with high (light) grayscale values, or "black hot," where the hottest objects in the scene are depicted with low (dark) grayscale values. Choice of hot-white or hot-black would be made by those performing interpretation of the imagery, depending on what is of particular interest in the scene.
FMV sensor packages are compact, portable packages that can be quickly installed on a wide variety of airborne vehicles, manned and unmanned. Specifications follow for many of the primary motion imagery platforms, both manned and unmanned. Each of these are primarily focused on intelligence gathering activities where extended loitering over areas of interest is of key importance to operations. Around the world, these aircraft can also be used to collect against dire environmental emergency events.
There are differences between classical remote sensing and FMV, and the skill set required for an analyst in these respective domains. In classical image analysis, success depends on the ability of the analyst to reliably identify specific objects of interest based on shape, texture, or radiometric signature. In FMV, success is less dependent on object identification and largely achieved through the ability of the analyst to work through shortcomings in the system architecture without losing the tempo required to maintain real-time operations. Whereas the classical image analyst may be the type of person who is very detail-oriented, thorough, and focused in a specific niche of expertise, the FMV analyst must be an interactive integrator of many sources of intelligence, capable of acting and making critical decisions without the aid of extensive research.
The power of FMV is brought to bear when the team controlling acquisition and exploitation of real-time video know precisely where and when to look at a target of interest. Discovering a target is not the goal of FMV. The goal is to follow a target in order to ascertain patterns of behavior which can then be used for a tactical advantage. In classical remote sensing, on the other hand, previously unknown or unsuspected information can often be discovered by detecting changes in a region of interest over time. These two disciplines can potentially complement each other, but the skills, training, and technology needed to support them are clearly quite different.