An overview of the Joint Service Imagery Processing System.

The Joint Service Imagery Processing System -- JSIPS, or the AN/TSQ-166V -- is a modular, tactically deployable system for receiving, processing, exploiting and disseminating digital imagery from multiple sensor types and sources. It also provides intelligence reports to the field commander in an average of 15 minutes from the receipt of real-time imagery. The system is being built in different configurations using the same basic components for the USAF, USMC, USA and USN.

Each JSIPS configuration consists of up to six segments:

* Tactical Input Segment

* Softcopy Exploitation Segment

* Hardcopy Exploitation Segment

* Exploitation Support Segment

* Communications Support Segment

* System Support Segment

These segments are functional divisions of the system, but relate loosely to the physical divisions. The system is tailored to meet specific user-unique requirements and operational capabilities by reconfiguring and/or repackaging the segments.

Most JSIPS systems are packaged in 8 x 8 x 10-ft or 8 x 8 x 20-ft shelters that are transportable by truck, rail, ship, aircraft or helicopter. The JSIPS full-scale development program implemented one configuration for the US Army in 8 x 8 x 20-ft shelters conforming to International Standards Organization (ISO) standards and a combined configuration for the USMC and USAF in 8 x 8 x 10-ft ISO shelters.

The JSIPS design provides for preplanned product improvements (|P.sup.3~I) via plug-in hardware or software modules. Examples of specific |P.sup.3~I requirements include SAR processing, automated target detection and identification, multisource correlation, weather data processing and remote operation of platform sensors from the SES.

The overall flow in a reconnaissance mission begins with the receipt of intelligence requests at a command center (an Air Control Center for the USAF). These requests are merged, sorted, prioritized, turned into requests for specific imagery and assigned to reconnaissance units. The units then assign specific aircraft and sensors to perform the collections and a specific JSIPS to receive the imagery transmission. Meanwhile, the exploitation requirements are sent to JSIPS so that the required intelligence can be extracted from the received imagery and reports can be written and sent to the requestors.

Within JSIPS, this flow can be traced through the system using Figure 1. An exploitation request from the command center enters JSIPS via communications lines into the CSS, through cryptographic decoding equipment and into the communications support processor. Requests are then entered into the computer-aided tactical information system (CATIS) data base in the ESS computer. CATIS then sends tactical tasking messages to the SES system management software, which forwards appropriate tasking to the TIS and the SES workstations. The TIS then enters the appropriate information regarding expected time and airspace coordinates for the aircraft's arrival into a queue of similar information for outstanding missions.

Once JSIPS starts reception from a platform, the imagery is received by the TIS, reformatted and sent to the SES. The SES contains a large array of disks that provides a rate-control buffer for the imagery before sending it to the workstations. A reduced resolution of the imagery "waterfalls" across the workstation monitor, where the interpreter can choose selected image target areas (SITAs) using an LED touch screen mounted in the CRT's bezel. These SITAs are saved for later detailed analysis to recover the intelligence needed for distribution to the original requestors.

KEY JSIPS HARDWARE TECHNOLOGIES

Imagery processing systems need large storage capacities and very fast transfer rates. Since JSIPS is packaged in field-deployable shelters, space is obviously at a premium. Several approaches were taken to reduce the amount of data, maximize transfer rates, and maximize storage densities in terms of bits/|in.sup.3~ (including the media, transport and associated electronics). Key technologies included near-lossless, high-density image compression; high-speed imagery and data LANs; and a storage hierarchy with the highest bits/|in.sup.3~ densities possible.

Minimization of storage capacities and transfer rates was accomplished through the use of an adaptive discrete cosine transform (DCT) compression of the imagery data into a JSIPS common image format at 1.3 bits per pixel (bpp).

High transfer rates between major equipment were achieved using an 80-Mb/s token-ring, fiber-optic, image LAN. Fiber optics also solved the EMI/TEMPEST requirements on emanations from signals passing between TEMPEST shelters. High transfer rates within major equipment were achieved through the use of high-speed, multiport, random-access memories.

Storage densities and access rates were maximized using the following:

* high-density, helically scanned, 19-mm, D-1 magnetic tape cassettes operating in modified Mil-Std-2179 tape cassette drives

* clusters of 760-MB Winchester disk drives paralleled to provide peak transfer rates of 75 Mb/s.

Other key technologies were shock isolation to meet stringent transportability requirements and a security design that allows the system to operate in a classified mode.

JSIPS Common Format

In JSIPS, image data derived from several sensor types (EO, IR and SAR) are put into a JSIPS common format. To create this format, the input data are first divided into 1,024 x 1,024 pixel tiles of imagery. Reduced resolution image set (RRIS) tiles (pyramidal image sets) are created from four adjacent image tiles using 2:1 decimation and filtering. These RRISs are used at the softcopy workstation for zooming (magnification and minification) of the imagery. The uncompressed original and reduced resolution imagery tiles are compressed to approximately 1.3 bpp using an adaptive DCT with 32 x 32 pixel transform blocks and Huffman encoding.

Off-Line Data Storage

The off-line data storage in JSIPS is implemented with helical-scan D-1 tape cassette recorders. D-1 tape cassettes are a family of 19-mm tape cassettes whose physical dimensions and tape characteristics have been standardized. The format of digital data recorded on D-1 cassettes can be controlled by either of two standards: Mil-Std-2179 or ANSI X3.175-1990.

The tape cassette recorder/players (TCRs) used on JSIPS are Mil-Std-2179 recorders, modified for azimuthal recording of randomized data. This modification improves the tracking and signal-to-noise margin, allows for a larger temperature operating range and requires less special tape handling. It also eliminates non-azimuth recording's potential for large signal-to-noise ratio losses and sync losses because of mistracking.

TACTICAL INPUT SEGMENT

The Tactical Input Segment (TIS) is the segment that interfaces with airborne sensor platforms, processes digital sensor data and provides processed and formatted imagery and support data for softcopy exploitation. The currently defined platforms with which the TIS will interface are the TR-1, U-2R, RF4-C, F-16R, F/A-18D/E and unmanned airborne reconnaissance vehicles. The TIS will accept three types of digital sensor data: electro-optic (EO) imagery, infrared (IR) imagery and, as |P.sup.3~I, synthetic aperture radar (SAR) phase-history data.

As shown in Figure 1, input to the TIS is via either couriered D-1 tape cassette or data link, in which case the data is recorded on a Mil-Std-2179 mission tape cassette recorder. Image data are processed in the TIS at rates up to 35 megapixels per second (Mp/s) by either an EO/IR processor or a SAR processor (|P.sup.3~I) and formatted into 1,024 x 1,024 pixel tiles of uncompressed imagery. RRISs are then created from the image tiles. The original imagery and the RRISs are compressed into the JSIPS common format, archived to a D-1 archive tape cassette and sent to the SES for softcopy exploitation over the image LAN.

The TIS operator interfaces with a TIS control computer. The software and TIS operator provide TIS control and management, including scheduling of processing, data-link capture and archive playback activities, tape management, image quality monitoring and interface to the SES via the Ethernet data LAN. In addition, the TIS operator has secure communication capability with the aircraft via UHF/VHF or HF radio with appropriate cryptographic equipment. The radios used in the TIS are the AN/ARC-182, multiband VHF/UHF radio and the AN/ARC-174 HF radio.

Primary Data Link

The JSIPS primary data link is based on the modular interoperable surface terminal (MIST). Two racks of equipment -- the link equipment group and the operator equipment group -- are housed in the TIS shelter. The antenna equipment group (consisting of a 6-ft parabolic dish and RF assembly) and an enclosure equipment group (communicates between the link equipment in the shelter and the antenna) are housed on a modified M794 flat-bed trailer, capable of being remoted up to 5 km using fiber optics. The antenna trailer is geolocated during set up of the system using the portable global positioning system receiver from the CSS or by manual means.

The JSIPS primary data link is designed to operate with airborne datalinks based on the modular interoperable data link (MIDL). The MIST/MIDL line-of-sight data link operates either in X- or Ku-band, tunable in 5-MHz steps, with down-link data rates of 10.71 Mb/s, 137 Mb/s and 274 Mb/s and a 200-kb/s up-link. Both the up-link and the down-link are composed of time-division multiplexed channels.

EO/IR Processing

The EO/IR processor accepts either compressed or uncompressed EO or IR imagery from up to two sensors simultaneously; synchronizes and, if necessary, decompresses the input; applies various pixel corrections if required; and returns the data to the TIS formatter, formatting the data into 1,024 x 1,024 tiles of uncompressed imagery. The EO/IR processor accommodates sensors with 6 to 12 bits per pixel, 512 to 15,000 pixels per line and up to 2,600 lines/sec (but with a pixel rate not to exceed 35 Mp/s). Although the EO/IR processor can accommodate three decompression schemes, the only one currently implemented is a 2- or 4-bit DPCM compression used on the Advanced Tactical Air Reconnaissance System (ATARS) program.

SOFTCOPY EXPLOITATION SEGMENT

The Softcopy Exploitation Segment (SES) can consist of up to six sets of shared equipment (computer, video printer, formatter, disk drive clusters and a Mil-Std-2179 tape drive) and up to 18 softcopy workstations, depending upon the configuration.

Although the SES formatter in each set of shared equipment can talk to any softcopy workstation, packaging constraints of the ISO shelters provide for one set of shared equipment and three workstations for the 10-ft ISO shelters and one set of shared equipment and six workstations for the 20-ft ISO shelters. Three side-by-side softcopy workstations, along one of the SES's 8-ft x 8-ft walls, are shown in Figure 4.

SES Softcopy Workstation

The softcopy workstation is based on a commercial-off-the-shelf (COTS) Unix CPU with graphics processor. The workstation is enhanced with a set of JSIPS boards for performing very fast image manipulations and enhancements. Each workstation has dual 1,280 x 1,024 pixels displays, mounted one above the other as shown in Figure 4, with a stereo display capability and a touch screen on the lower monitor.

Image manipulations include a 350-pixel/sec visually continuous roam, a zoom from 1/32x to 16x and a rotation (|Mathematical Expression Omitted~ minimum increment). The display is updated every 1/60th of a second for roam and 4/30th of a second for zoom and rotation. All image manipulations and enhancements can be performed in any order. For example, a warped image with the dynamic range adjusted and edge enhanced can be rotated, zoomed and roamed from one edge of the image to the other without pausing at RRIS boundaries or for recalculation of warping coefficients.

Each softcopy workstation receives imagery via the 80-Mb/s image LAN and airborne platform support data (e.g., navigation data, sensor pointing data) via the Ethernet data LAN. Each workstation has a 44-MB removable media disk drive, a 600-MB CD-ROM disk drive for the input of mapping data and two 760-MB disk drives, one for imagery and one for use as a system disk. The human interface is provided by a ruggedized keyboard, trackball, image enhancement and rotation knobs, rocker switch for zooming and joystick for translation (roaming).

Workstation Operations

The workstation operations include screening, detailing (exploitation, report generation and the creation of secondary images for dissemination outside of JSIPS) and the review and validation (R&V) of images to be disseminated. (System supervisor, operations supervisor, system security officer or maintenance operator functions can also be performed, depending upon user log-on.)

Screening is the designation of selected image target areas from the input imagery that require detailed interpretation, discarding all input imagery that does not contain targets of interest. Such purging of imagery is important in managing the quantities of data in an image processing system. Screening can be either automatic, auto-assist or manual.

The detailing operator performs exploitation, report generation, and creation of secondary images for dissemination outside of JSIPS. A list of imagery prestaged to the workstation is presented to the detailer, who then displays the image, along with a display showing details of the mission and the platform and sensor positions corresponding to the center of the image. In other windows, the detailer can also display reference imagery. Pull-down menus are used to select prompt boxes for additional information or command entries displayed on the right side of the screen, with minimal, if any, overlap of the imagery.

In performing the exploitation, the operator uses a variety of workstation capabilities to extract the needed intelligence from the imagery. These capabilities include:

* image manipulation: roam, zoom, rotation

* image enhancement: dynamic range adjustment, edge enhancement, smoothing/noise filtering, warping.

* image display: pseudo-color, brightness, density slicing, registration with a reference image, addition/subtraction of two images, flicker between two (registered) images, stereo, overlay of graphic cues (SIGINT error ellipses, target symbols), image windows, split screen and quad screen.

Additional capabilities used in exploiting the imagery are target location (geopositioning) and object measurement (e.g., length, area, height, volume) based on cursor-defined locations. Four target-location techniques are available, all of which use collection platform navigation data, sensor-taking parameters and a sensor geometry model.

Report generation is performed on the softcopy workstation by using a window for CATIS workstation emulation. The detailer accesses a partially prefilled report from the CATIS data base containing entries such as target type, location and security classification of the report and completes the report with information such as target description and significance, weather, imaging conditions and free text descriptions.

Secondary imagery is created at the workstation. After the image is selected for dissemination, it must be reviewed and validated for classification level by an R&V operator, either coincidentally with or subsequent to the detailer's selection of the image for release. After review, the image to be disseminated is queued for release by the image release host in the Communications Support Segment.

HARDCOPY EXPLOITATION SEGMENT

The Hardcopy Exploitation Segment (HES) provides the US Army the means to:

* perform softcopy-to-hardcopy conversion (create film)

* perform hardcopy-to-softcopy conversion (digitize hardcopy)

* perform image interpretation at a light table

* perform precise target location using an APPS-IV

* produce reports to be exported to JSIPS customers.

Softcopy-to-hardcopy conversion is provided using an electron-beam hardcopy reconstruction unit (HRU) and front-end processor. The front-end processor receives imagery from the SES in JSIPS common format, decompresses and buffers the imagery, translates the support data into titling information and transmits the imagery and associated titling to the HRU for printing onto film.

Hardcopy-to-softcopy conversion is provided with a 300-dots/in. digitizer. The digitized imagery is transferred to the SES via either 8-mm tape cassette of the Ethernet data LAN.

Hardcopy exploitation is performed using film on light table workstations with stereo optics. Light table mensuration and geopositioning is provided by connection to a ruggedized HES computer, which is also used for overall HES control. Three-dimensional geometric models used for mensuration and geopositioning reside on the HES control computer and use coordinates input from the light table.

Precise target location is performed using an APPS-IV. The APPS-IV uses a stereo viewer with stereo film pairs from the geographically referenced, hardcopy, point positioning data base (PPDB) distributed by the Defense Mapping Agency.

Reports are composed using terminals that connect via the Ethernet data LAN to the ESS computer. These terminals provide windows into the reporting capability provided by the ESS's CATIS software.

EXPLOITATION SUPPORT SEGMENT

A central component of the Exploitation Support Segment (ESS) is a data-base management system that contains geolocations for all exploitation requirements, collected imagery and auxiliary data (such as target lists) in the system. Using this geographical data base, the ESS provides:

* requirements and dissemination management

* hardcopy and softcopy exploitation management

* imagery interpretation report creation and transmission

* incoming report processing

* data-base management

* exploitation-report processing.

The ESS functions are accomplished by the Computer-Aided Tactical Information System (CATIS) software and the JSIPS software augmentations required to make it a tactically useful system.

Exploitation reports output from JSIPS include the initial, supplemental and multimission imagery photo interpretation reports and the reconnaissance exploitation report. Completed reports are queued for review and validation by a supervisor and then for release external to JSIPS.

COMMUNICATIONS SUPPORT SEGMENT

The Communications Support Segment (CSS) provides communications both within JSIPS and between JSIPS and the outside world. These communications include voice, messages and imagery. In addition, the CSS includes a GPS receiver that is used for providing both a precision time source for the system and a precise location for the TIS data-link antenna during system set-up.

JSIPS voice communications consist of intercom instruments and an intercom private branch exchange (PBX) for internal communications; secure telephone units (STU-IIIs) and a telephone PBX for connection externally either to a base telephone exchange (and hence to the public switched telephone network) or to the military's automatic voice network (AUTOVON); a separate STU-III for connection to the military's automatic secure voice network; and KY-68 digital secure voice terminals for connectivity to certain common user and dedicated user lines when the use of other external communications is not possible during deployed operations.

Message communications to and from JSIPS are through the military's Automatic Digital Network (AUTODIN) (Mode I). Since JSIPS can operate at a classified level, an automated message processor exchange is required to connect JSIPS to AUTODIN. JSIPS uses the accredited, multi-level security communications support processor (CSP) to fulfill this requirement. Connections from the JSIPS CSP are also provided to the enhanced tactical user's terminal and electronic processing and dissemination system for some JSIPS configurations.

Imagery communication through the CSS is via an imagery release interface (IRI), enabling the system to exchange secondary imagery with the Intratheater Imagery Transmission System, tactical digital facsimile, fleet imagery support terminal/imagery support terminal and National Imagery Transmission Format systems. The IRI communicates using the Defense Data Network, Tri-services Tactical Communications, the public switched telephone network or dedicated lines.

SYSTEM SUPPORT SEGMENT

The System Support Segment (SSS) houses, powers and cools the other five JSIPS segments. The JSIPS 8 x 8 x 10-ft and 8 x 8 x 20-ft shelters are based on Mil-Std-906. They are modified to include the equipment racks, power filters, red/black cable separation and duct openings for the environmental control units. The SSS is subdivided into the shelter support group, a mobile power group, an environmental support group, a hardcopy digitizer group and a maintenance data computer.

In addition to the TIS, SES, HES and CSS/ESS shelters, another shelter is provided as part of the SSS for storage of spares and maintenance equipment and for a maintenance area, including space for a maintenance data computer used for inventory control and record keeping.

SUMMARY

JSIPS is the first fielded modular, tactically deployable system capable of receiving, processing, exploiting and disseminating digital imagery from multiple sensor types and sources. The system has been required to undergo extensive product assurance tests, including reliability verification, environmental stress screening, environmental testing (temperature/humidity/altitude), transportability and mobility testing and electromagnetic compatibility testing.

Key JSIPS technologies include near-lossless image compression, high-speed LANs, a storage hierarchy with high bits/|in.sup.3~ densities, shock isolation to meet stringent transportability requirements and a security design to allow system operation in a classified environment connected to communication links operating at lower classification levels. Soon field commanders will have the access they need to current, in-depth tactical images and reports.

Robert Davis received a BSEE from the University of Illinois in 1965, MSEE from Georgia Institute of Technology in 1967 and Ph.D in electrical engineering from the University of Pennsylvania in 1979 in the area of speech and pattern recognition. He has worked at Burroughs (now Unisys), Texas Instruments' Central Research Laboratories, SD-Systems and a European joint venture between Honeywell and L.M. Ericsson. Since 1985, Dr. Davis has been with E-Systems working on JSIPS and other imagery exploitation new business pursuits. He was a visiting professor at Southern Methodist University from 1983 through 1988 and is a senior member of IEEE and a member of SPIE.

Article Details

Printer friendly Cite/link Email Feedback
Author:	Davis, Robert L.
Publication:	Journal of Electronic Defense
Date:	Dec 1, 1992
Words:	3420
Previous Article:	Reaping the RF harvest.
Next Article:	Trends in electronic counter-countermeasures.
Topics:	Digital electronics Evaluation Imaging systems Evaluation Military electronics Evaluation