After an outline of fundamentals of the multilens array combined with focal plane shutter and angle-correct FMC the KA-106() and KA-107C will be described. Then the special ground resolution capability of the multilens camera compared to that of the tri-camera fan and the geometric imaging capability compared to that of the panoramic scan camera will be discussed.

The model 700 camera is the latest in a 30-year series of LOROP cameras developed by McDonnell Douglas Astronautics Company (MDAC) and their predecessor companies. The design achieves minimum size and weight and is optimized for low-contrast performance. The optical system includes a 66-inch focal length, f/5.6, apochromatic lens and three folding mirrors imaging on a 4.5-inch square format. A three-axis active stabilization system provides the capability for long exposure time and, hence, fine grain films can be used. The optical path forms a figure "4" behind the lens. In front of the lens is a 45° pointing mirror. This folded configuration contributed greatly to the lightweight and compact design. This sequential autocycle frame camera has three modes of operation with one, two, and three step positions to provide a choice of swath widths within the range of lateral coverage. The magazine/shutter assembly rotates in relationship with the pointing mirror and aircraft drift angle to maintain film format alignment with the flight path. The entire camera is angular rate stabilized in roll, pitch, and yaw. It also employs a lightweight, electro-magnetically damped, low-natural-frequency spring suspension for passive isolation from aircraft vibration inputs. The combined film transport and forward motion compensation (FMC) mechanism, which is operated by a single motor, is contained in a magazine that can, depending on accessibility which is installation dependent, be changed in flight. The design also stresses thermal control, focus control, structural stiffness, and maintainability. The camera is operated from a remote control panel. This paper describes the leading particulars and features of the camera as related to weight and configuration.

Technological and operational requirements for airborne photographic reconnaissance missions over the past several years have generated the need for a multi-mission, multi-mode sensor control system. It is desirable that the sensor control system maximize operator-pilot mission flexibility while minimizing operator-pilot interaction with the system and the sensors under its control. Ideally the sensor control system would be capable of performing its role when install-ed in a tactical fighter aircraft, which is subject to environ-mental and operational extremes, with little or no degredation of reconnaissance mission performance. This paper describes an integrated sensor control system developed by the Fairchild Space & Electronics Company (Fairchild) for the Northrop Corporation for use on the RF-5E reconnaissance aircraft.

A road-finding algorithm has been developed which utilizes a new method for detecting straight edges in a scene. The road finder has accurately identified roads in a variety of images such as infra-red and multispectral, forward and down-looking, and rural and urban environments. The algorithm is designed to locate roads that are between three and twelve pixels in width. The new edge detector takes advantage of the unique characteristics of the edge-preserving smoothing filter of Nagao and Matsuyama, which has been improved by using it in combination with median filters. Data reduction is guided by a knowledge of road characteristics - edges in the scene are efficiently reduced to instances of anti-parallel segments. Combinations of these segments are then evaluated with respect to several road scoring functions to determine the location and confidence of the best road in an image.

Fisher. On my left is Robert E. Tracy. Bob probably has the honor of being a senior member on this panel and he has been involved in reconnaissance, one way or another, since about 1939. He has some real first-hand experience in that he actually flew as an aerial photographer in B-47s and RB-36s. He's also a qualified image interpreter. Since 1969 he has been project engineer on one of our advanced strategic aircraft. He earned his BS degree at Southern Illinois University and his MSA at Cal State. He has also pursued advanced studies in recce intelligence and various CAI, DIA, USA and industrial schools. He's a certified photogrammetrist, and a co-inventor on a patent for a correlation system for multi-sensor re-connaissance vehicles. Also, he's a member of several national groups dealing with reconnaissance and intelligence. I have been told by some people who have known him for a long time this his approach doesn't always follow the mainline thinking. So he fits in real well with this group.

Until recently the development of FLIR Automatic Target Recognizers (ATRs) has been constrained by several pitfalls, including an inadequate specification and choice of training data. Moreover, the best data bases in the world have proven to be of little value when surrogate features are extracted without scientific cunning. Ad hoc selection of the classifier is the final pitfall.

Several automated procedures have been developed and tested on photo interpretation students. Although these techniques are certainly more reliable and accurate than by-hand work, this was not the primary purpose behind their development. The goal, in fact, was to design photo analysis procedures which were more rapid and less subject to blunders. Two specific areas of development have been investigated. The first involves the use of the digitizer and computer for oblique mensuration work. Several self-documenting programs have been written to allow the photo interpreter to quickly and accurately determine the depression angle, then measure various kinds of images. The second kind of tool involves the use of an image analyzer in complex enumeration tasks on aerial imagery. This technology has been applied commonly in biology and elsewhere, but is not usual in the aerial photo interpretation process.

The feasibility of real-time imagery interpretation is examined from the perspective of the operator in the loop. The question of whether or not military analysts are capable of performing adequately in present high and forthcoming higher throughput environments is addressed. In an attempt to shed light on this situation, two radar imagery exploitation systems, the Flexible Test Bed (FTB) and Advanced Building Blocks for Large Area Exploitation (ABLE), are reviewed. Softcopy imagery exploitation performance achieved during operational field training exercises in the Federal Republic of Germany is presented.

The airborne system for aerial reconnaissance (SARA is the outcome of the Air Defence OHQ's intention to acquire a complete system for replacing the MIRAGE III R and RD in operation since 1963.

The proliferation of a number of similar image processing ground stations has resulted in an array of image processing hardware with diverse logistics, training, and reporting requirements. Contemporary system approaches lack the flexibility to accommodate changing sensor suites and to meet future intelligence timeline requirements. System concepts are currently being evolved under ongoing study efforts, with the objective of resolving these issues. As discussed in this paper, architectures are being considered which are reconfigurable and which will have the flexibility for processing a wide range of sensor types including Synthetic Aperture Radar (SAR), Infrared (IR), and Visual. Through commonality and modularity, a common base of hardware/software will satisfy a multitude of image exploitation needs, and as a result, should be much more cost-effective and simpler to support logistically. The application of advanced image processing techniques, including automatic target classificaton, pattern recognition, and automatic point position data base retrieval techniques, will offer significant improvements in system manning requirements, throughput, target location accuracy, and result in more consistent, expedient target reporting. Utilization of VLSI technology for processor implementation will result in major size reductions, thus improving transportability for rapid deployment.

Duplication of high-quality continuous-tone photography requires a capability to accept broad ranges of contrast and maximum density in the original negative. An electrophoto-graphic system using liquid toner and a final-imaging photoconductive film has been shown to meet these requirements. The system can be applied in continuous roll duplication of high-quality originals. A description is given of the special characteristics and control of the photoconductive film that make it applicable in this system, taking advantage of the property of persistent conductivity to augment conventional electrostatic image formation.

The Canadair CL-227 is a rotary winged Remotely Piloted Vehicle (RPV) intended initially as the air-vehicle for a medium range battlefield surveillance and target acquisition system. The concept on which this vehicle is based brings together in-house expertise as a designer and manufacturer of surveillance drones (AN-USD-50l -MIDGE-) with experience in rigid rotor technology from the CL-84 tilt wing VTOL program. The vehicle is essentially modular in design with a power module containing the engine, fuel and related systems, a rotor module containing the two counter-rotating rotors and control actuators, and a control module containing the autopilot, data link and sensor system. The vehicle is a true RPV (as opposed to a drone) as it is flown in real time by an operator on the ground and requires relatively little skill to pilot.

The AN/USD-501 and AN/USD-502 System airborne vehicles are reusable, high subsonic speed, missile-shaped drones. Primary difference between the two drones are in the flight range capability, flight control and navigation. sophistication, and payload capacity and capability, The AN/USD-502 drone contains a single sensor. The AN/USD-502 drone contains one large payload compartment for either a single large sensor or multiple smaller sensors.

Infrared Oblique reconnaissance from an aircraft can be achieved either by use of FLIR (Forward Looking Infrared) sensors or by using IRLS (infrared line scanners) when such sensors are pointed in an oblique direction. Each sensor has certain problems associated with its use. The FLIRs offer good resolution and sensitivity but have a narrow field of view while the IRLS sensors have wide coverage but suffer from sampling overlap and underlap mismatch in each scan. These problems are discussed. Some solutions are mentioned which tend to maximize resolution and sensitivity so that oblique imaging at long slant ranges near the horizon can be accomplished on a low altitude reconnaissance mission. Current development activity at Honeywell toward these goals is briefly noted.

Sophisticated reconnaissance systems often must trade reliability for high operational performance. The resulting increased maintenance requirements dictate a more comprehensive maintainability program to insure system availability. To improve efficiency and lessen reliance on human performance, advanced maintainability concepts must be implemented beginning at design onset to insure operational success after system deployment.

This paper reports on the development and tests of a new stable film developer that can fully exploit all the inherent speed in the newer thin emulsion, high resolution, high contrast films such as Kodak Technical Pan and Aerographic II, and do this at a low contrast and in the short development times necessary for routine large tank processing. The concept exploits the super additivity of phendione-hydroquinone to push process the "toe" for shadow details. The contrast is held down by reducing the flow of replacement developer through the emulsion to the highlight portions of the negative and by controlling the activity of the hydroquinone by alkaline buffering. Effective exposure index (EEI) is greatly boosted over ASA or A FS speeds by placing an early bump in the D-Log E curve.

Experiments carried out over diagonal lines of sight through the entire atmosphere support the concept of spatial coherence degradation through forward scattering as described by an aerosol transfer function which strongly affects the wavelength dependence of imaging through the atmosphere. Airborne-particulate size and concentration are affected strongly by wind strength and soil moisture. Changes in weather that result in changes in average particulate size of airborne soil-derived particulates also strongly change the wavelength dependence of resolution through the atmosphere as a result of changes in the wavelength dependence of the scattering coefficient. Knowledge of such effects can therefore permit prediction of spectral regions most suitable for imaging through the atmosphere.

Alignment errors among components of an optical system may substantially degrade the image quality. Focus errors also affect system performance. The potential for serious degradation of image quality is substantial and requires that the tolerances for these errors receive significant attention early in system design. The image quality and reconnaissance performance of an all-reflecting Cassegrain is compared to an all-refractive optical system under conditions of zero and anticipated "real world" misalignments.