SPIE Membership Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:
SPIE Photonics West 2018 | Call for Papers

SPIE Defense + Commercial Sensing 2018 | Call for Papers




Print PageEmail PageView PDF

Defense & Security

Air-to-ground image pixels mapped to locate hostile events

Advances in surveillance software that maps sensor pixels to ground coordinates allow the study of diverse scenarios for the detection of hostile events and the tracking of their perpetrators.
7 August 2006, SPIE Newsroom. DOI: 10.1117/2.1200607.0320

In the general context of the war on terror, automated aerial surveillance, detection of hostile ground events and tracking of perpetrators are considered critical factors in the prevention and control of insurgent uprisings. This type of warfare increasingly depends upon systems and services external to weapon systems, referred to as Command, Control, Communication, Computer, Information, Surveillance and Reconnaissance (C4ISR) systems. To be effective, C4ISR technology requires precise pixel (row,column) mapping into ground location (latitude, longitude, altitude) given platform location and orientation.

Progress in this field requires consideration of several C4ISR footprint tradeoffs, which include: the number of sensors, the individual sensor field of view (FOV), the variability of the geographical area covered as a function of platform position and attitude, and ground surface elevation variations and uncertainties. Event pixel location specification includes latitude, longitude, altitude for the pixel centroid and corners, and line-of-sight (LOS) centroid range. In addition, these basic requirements are further constrained by the sensor pixel count, optics and their related distortions, platform pointing stability, data handling and storage limitations, and system cost and platform load limitations.

Further, from high altitudes, sensor pixels can only map to rectangles on the ground with the result that only target tops are seen. At lower altitudes, perspective distortion and the effect of platform attitude variation considerably reshape pixel footprints. Additional pixel distortions arise from the earth's oblate spheroid shape, which flattens toward the poles. This produces line-of-sight far field ranges that, at mid-latitudes, vary significantly with compass direction. Finally, urban footprints, affected by the LOS obscuration due to the presence of buildings, require special data handling techniques to cope with the density of data.

To illustrate the mapping of surveillance system sensor pixels to the ground, we present modeling results obtained for two applications. The first is a 360°-viewing focal-plane-array sensor system for low altitude aerostats, and the second, a fixed area of constant surveillance for high altitude fixed-wing aircraft. In our studies, we model the earth as a sphere with a radius fixed by the platform latitude.

Figure 1 shows the footprint for four focal-plane arrays (FPA) of sensors, each with a 43° depression angle, mounted on an aerostat at an altitude of 1.3km. In Figure 1(a), the 30km red circle defines the area of required coverage. The 43° angle allows almost constant ground coverage of this area for most realistic aerostat attitude variations. Figure 1(b) shows a hypothetical event, declared by the north facing sensor, at row #100 and column #200 (one o'clock) at a ∼24km radius. In Figure 1(c), the black pixel corners and red centroid are specified in latitude and longitude using an earth radius of 6371km.

Figure 1. (a) Shown is the footprint for four 320×240 pixel FPAs (focal plane arrays) modeled for an aerostat platform at an altitude of 1.3km (39°N,77°W); (b) target pixel (at one o'clock) on the flat-earth; (c) the pixel centroid (red) and corners (black) are projected onto a spherical earth and specified in (latitude,longitude) coordinates.

Figure 2 shows a fixed-wing aircraft scenario in which six sensors provide a 14km diameter region of constant surveillance (ROCS) at a pixel resolution <1m2. Figure 2(a) shows the geometry and specifications for two rows of sensors, each having different FOVs, elevation angles, and sensor spacings. The position of the aircraft is the red dot at the center of the range concentric circles. As the airplane circles, the footprint rotates about the center of the ROCS, shown as a black dot in Figure 2(a). The sensor footprint outside the black dashed circle is not part of the ROCS. For this scenario, we modeled each sensor to have a square FOV with 16,000 pixels on a side. The maximum pixel size within the ROCS, shown in Figure 2(b), is approximately 1m2 at a LOS range of ∼23.3km.

Figure 2. (a) Schematic of six beams directed from a fixed wing airplane (red dot) circling a region of constant surveillance (dashed black circle). Designed for a platform altitude of 6km, the six beams as illustrated produce a ROCS diameter of 16km at a 14km flight path radius; (b) the pixel pattern near the overlap of the red and blue beams shows 1m2 resolution at a LOS range of 23km.

We have developed fieldable surveillance software that maps sensor pixels to earth coordinates and allows testing variations in air-to-ground surveillance scenarios. Our next step will include more precise earth models and high density building elevation data. This will allow us to address hostile event detection and tracking within urban areas. Additional studies will be required to include estimations of the effect of navigation errors on pixel location specification and the effect of platform pointing instabilities on ROCS duty cycles.

Bruce Weber, Joseph Penn
AMSRD_ALC_SE_SE, U. S. Army Research Laboratory
Adelphi, MD
Dr Weber is a senior research physicist. His current research interests include surveillance system development, imaging and non-imaging sensor automatic target recognition, and synthetic target signature development and validation.
Joseph Penn is a research physicist whose research interests include surveillance system design and development, and synthetic target signature development.