WO1997042765A1

WO1997042765A1 - High resolution and wedge-filter camera system for low earth orbit satellite imaging

Info

Publication number: WO1997042765A1
Application number: PCT/US1997/007814
Authority: WO
Inventors: Arno G. Ledebuhr; Joseph F. Kordas; Preston H. Ii Carter; Warren T. White
Original assignee: The Regents Of The University Of California
Priority date: 1996-05-06
Filing date: 1997-05-06
Publication date: 1997-11-13
Also published as: AU3061897A

Abstract

A small satellite imaging system comprises a very high resolution (VHR) camera (20) fastened to a mounting plate such that its central line of sight is approximately perpendicular to the plane of the mounting plate, and comprising a beryllium telescope coupled to a silicon charge coupled device (CCD) responsive to the visible and ultraviolet spectrums. A wedge-filter camera (18) is fastened to the mounting plate such that its central line of sight is also approximately perpendicular to the plane of the mounting plate, and comprising a beryllium telescope coupled to a silicon charge coupled device responsive to the far and near infrared spectrums with an interposed linear variable filter that passes light wavelengths to the CCD depending on the spatial orientation of the source to the central line of sight.

Description

HIGH RESOLUTION AND WEDGE-FILTER CAMERA SYSTEM FOR LOW EARTH ORBIT SATELLITE IMAGING

The United States Government has rights in this invention pursuant to Contract No. watts-7405-ENG-48 between the United States Department of Energy and the University of California for the operation of Lawrence Livermore National Laboratory.

COPENDING APPLICATION This application is a continuation-in-part (CIP) of United States Patent Application 08/522,183, filed 8/31/95, and incorporated herein by reference.

BACKGROUND OF THE INVENTION Field of the Invention

The present invention relates to camera and imaging technology and more particularly to multi-spectral imaging of the earth from small satellites in low orbits. Description of Related Art

Satellites present extraordinary platforms from which camera images can be collected. But the satellites themselves also present major challenges to the design, testing and operation of such cameras. The weight and size of the camera systems are critical. By one estimate, it costs $10,000 to orbit each pound of payload. So "small-satellite" technology is experiencing a comeback. The cameras that are launched must be able to resolve fine features from the very great distances involved in earth and moon orbits and transits. The orientations of the satellites themselves is also critical to aiming the cameras. The conventional applications for imaging and non- imaging spectrometers in space have been extensive. For example, twelve-channel prism and nine-channel grating spectrometers have been used for space-borne sensing of terrestrial resources. A thirteen-band multispectral scanner was flown on the Skylab and measured spectral bands in the range between 410 nanometers and 2350 nanometers, each band 20-100 nanometers wide.

The LANDSAT-D satellite used two scanning-type instruments, a thematic mapper and a multi-spectral scanner (MSS) which had four channels, 500-600 nanometers (green), 600-700 nanometers (yellow), 700-800 nanometers (red and near infrared) and 800-1100 nanometers (infrared). The Nimbus satellite included a coastal zone color scanner (CZCS) which used a grating spectrometer and five visible-near IR channels. The visible channels each had a spectral bandwidth of about 20 nanometers and were centered at 443 nanometers (blue), 520 nanometers (green), 550 nanometers (yellow), and 670 nanometers (red). The infrared channels had a 100 nanometers band in the near IR centered at 750 nanometers and a 2,000 nanometers band in the far IR centered at 11,500 nanometers. A dichroic beam splitter was used to separate the far IR radiation band from the visible radiation band. A moving mirror is used in a scanning system to view different parts of the scene across the array of detectors to collect multispectral images. Each detector operated at a different wave-band. A fourteen-channel radiometer, using combinations of detectors and filters was used on the earth radiation budget (ERB) sensor flown on the Nimbus-7 satellite. The solar back scatter ultraviolet (SBUV) used a moving-grate spectrometer to monitor twelve selected narrow wavelength bands. A filter photometer was used to measure a fixed band. The total ozone mapping spectrometer (TOMS) measured six discrete wavelengths in 1.0 nanometers bands. Both instruments have also flown on the Nimbus satellite.

The French SPOT satellite used two high resolution visible (HRV) imaging sensors. An included multispectral sensor used charge coupled device (CCD) arrays with filter-based spectral bands centered at 550 nanometers (green), 650 nanometers (red), and 840 nanometers (near IR). Each had an 80 nanometers bandpass. A panchromatic CCD used had a band-pass of 500-900 nanometers. NASA's multispectral linear array (MLA) used four fixed band CCD bandpass channels, 460-470 nanometers, 560-580 nanometers, 660-680 nanometers, 870-890 nanometers. Two near IR CCDs had fixed bands from 1230-1250 nanometers and 1540-1560 nanometers. In 1987, NASA placed in service the airborne visible infrared imaging spectrometer (AVIRIS). It was one of the most advanced imaging spectrometers of its time and used 244 bands, each with a 9.6 nanometers bandwidth. A new generation of imaging spectrometers is scheduled by NASA to be flown on-board the Space Station in the late 1990s, e.g., the high resolution imaging spectrometer (HIRIS), and the moderate-resolution imaging spectrometer (MODIS). Both use area arrays to obtain spectrally- resolved images of a one-dimensional scene. But, the spectral resolution is obtained with a diffraction grating having a 10 nanometers band-width. See: "Space-Based Remote Sensing of the

Earth", prepared by NOAA and NASA, 1987, "Nimbus 7 Users' Guide", published by NASA August 1978, "Remote Sensing of the Environment", by J. Lintz and D. S. Simonett, published by Addison-Wesley, 1976, "Earth Observing System-Instrument Panel Report", volumes lib, MODIS, and lie, HIRIS, published by NASA in 1987.

Wedge imaging spectrometers have been suggested for orbiting multispectral and hyperspectral acquisition systems, e.g., by George T. Elerding, et al., in "Wedge Imaging Spectrometer:

Application to drug and pollution law enforcement", SPIE Vol. 1479, Surveillance Technologies (1991). This work at Hughes Aircraft Company (Goleta, CA) credits the Santa Barbara Research Center (SBRC) and describes a sensor that is rugged and compact. To avoid the need for complex and delicate fore optics, a wedge filter is attached to an area array of detectors for two-dimensional sampling of the combined spatial/spectral information passed by the filter. United States Patent 4,957,371, issued September 18, 1990, to Samuel F. Pellicori, et al., also describes a wedge-filter spectrometer. Such Patent is incorporated herein by reference. The use of a continuous variable optical filter permanently aligned with a spectrometer sensor array in air-borne ground survey application is described by N. Gat, in United States Patent 5,166,755, issued 11/24/92, and incorporated herein by reference. A typical wedge filter has a transparent substrate on which several thin-film depositions are made with a taper in thickness along the y-axis of the plane of the substrate and no taper along the x-axis. The depositions result in light passband preferences that vary, e.g., from 400-1030 nanometers, in the wavelengths of light along the y-axis of the substrate plane that are allowed to pass according to the thickness of the thin-film depositions for each x-axis row. Thus, a rasterized CCD array, or equivalent, behind the wedge filter and an objective lens in front of the wedge filter will provide row-by-row slices of the spectrum, e.g., from 400-1030 nanometers for an entire raster from top to bottom. SUMMARY OF THE INVENTION An object of the present invention is to provide an imaging system that weighs only a few pounds and yet can analyze a broad spectrum of light and resolve terrestrial features as small as ten meters from orbit. A further object of the present invention is to provide a camera for multi-spectral imaging.

Another object of the present invention is to provide an imaging system for use in low earth orbit.

Briefly, a small-sat imaging system embodiment of the present invention comprises a very high resolution (VHR) camera fastened to a mounting plate such that its central line of sight is approximately perpendicular to the plane of the mounting plate, and comprising a beryllium telescope coupled to a silicon charge coupled device (CCD) responsive to the visible and ultraviolet spectrums. A wedge-filter camera is fastened to the mounting plate such that its central line of sight is also approximately perpendicular to the plane of the mounting plate, and comprising a beryllium telescope coupled to a silicon CCD responsive to the far and near infrared spectrums with an interposed linear variable filter that passes light wavelengths to the CCD depending on the spatial orientation of the source to the central line of sight. A satellite platform has the mounting plate attached and provides for aiming and ground communication with the VHR and wedge-filter cameras from earth orbit. It also provides for the orientation of the wedge-filter camera CCD relative to the direction of travel of the satellite platform in its earth orbit.

An advantage of the present invention is that a system is provided that is exceptionally light weight and rugged for its performance. Another advantage of the present invention is that an imaging system is provided for use in low earth orbit.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a block diagram of a small-sat imaging system embodiment of the present invention; Fig. 2 is a perspective view of the payload package for the satellite of Fig. 1 comprising the very high resolution camera and the wedge-filter camera;

Fig. 3 is a cross-sectional diagram of the very high resolution camera of Figs. 1 and 2; Fig. 4 is a function block diagram of the very high resolution camera of Figs. 1-3; and

Fig. 5 is a cross-sectional diagram of the wedge-filter camera of Figs. 1 and 2.

DETAILED DESCRIPTION OF THE INVENTION

Fig. 1 represents an imaging system embodiment of the present invention, referred to herein by the general reference numeral 10. The imaging system 10 includes a low earth orbit (LEO) satellite 12 with a power conditioning and status unit 14, a set of temperature sensors 16, a wedge-filter camera 18, a very high resolution (VHR) camera 20, a stepper-motor -operated filter wheel 21, a payload processor 22 and a solid-state memory 24. The general satellite functions include a low earth orbit experiment modulator 26 and an antenna 30 for transmitting out data on a 2490 MHz carrier at 2.5 Mbps. Because of the differences in magnification used between the cameras 18 and 20, the entire field-of-view of the VHR camera 20 spans only a small rectangular box, e.g., 19x14 pixels, in the field-of-view of the wedge filter camera 18. A satellite processor 32 handles general satellite command and status traffic and is connected to an antenna 34 for transmitting out data on a 2315 MHz carrier at 250K bits per second and a receiver antenna 36 for accepting commands on a 2070 MHz carrier at 9600 bits per second. A remote site control center (RSCC) 38 supplements a mission control center (MCC) and has a telemetry transceiver 40 with antennas 42, 44, and 46 for communicating with the satellite 12.

The telemetry is connected to mission planning 48, data analysis 50 and data storage 52. The planned experiment of October 1995 was expected to take the satellite 12 into a circular orbit around the earth at 400 kilometers with an orbital inclination of 40°±0.5°. In Fig. 2, a housing 60 provides a satellite payload package for the wedge camera 18 and the VHR camera 20. For example, the housing 60 is about fourteen inches tall (X), ten inches wide (Y), and fourteen inches deep (Z).

Referring to Fig. 3, the VHR camera 20 comprises diffraction-limited beryllium telescope optics including a primary mirror 70 and a secondary mirror 72 that directs images through a central bore 74 and a system of lenses 76 to a focal plane array (FPA) 78. A focusing adapter housing 79 connects the FPA 78 to the telescope optics. The satellite 12 is maneuvered to aim the VHR camera 20 appropriately. In the operational system prototype that was constructed, the FPA 78 was a Thomson-CSF type silicon CCD array, TH 7863 CRH-UV-01-B/T metrachrome-II, with a quantum efficiency (QE) > 8% at 250 nanometers. A special coating is preferably applied to increase the ultraviolet spectrum response. In Fig. 4, a camera 80, similar to VHR camera 20, functionally comprises a baffle 82 to exclude extraneous light, a lens 84 for focus, a six-color filter wheel 86 for band selection, a CCD 88 for conversion of the optical image to an electronic image with a focal plane array (FPA), a TTL to MOS level driver 90, a CCD controller 92, a SASI bus receiver 94, a unit 96 for setting sensor gain and offset, a flash analog-to-digital converter (ADC) 98, an electronics module 100 for amplification, filtering and double correlated sampling, a filter wheel controller 102, a SASI receiver 104 to latch commands, a power supply filter 106, a differential line driver and receiver 108, a voltage regulator 110, a lens temperature sensor 112, a CCD temperature sensor 114, a differential line driver and receiver 116, and a pair of a differential line drivers 118 and 120. The six-color filter wheel 86 is equivalent to filter wheel 21. The CCD 88 is equivalent to FPA 78. Referring to Fig. 5, the wedge-filter camera 18 comprises a pair of collection lenses 130 and 132, and a primary mirror 134 that direct images through a pair of focusing lenses 136 and 138 to a wedge-filter 140 and a focal plane array (FPA) 142. The satellite 12 is maneuvered to aim the wedge-filter camera 18 appropriately. In the prototype that was constructed, the FPA 112 was a Thomson-

CSF type silicon CCD array, TH 7863 F02-01-B/T fiber optic window, with a quantum efficiency (QE) of 40% at 650 nanometers. The bandpass of the wedge filter 140 typically depends on the wavelength involved, e.g., for 550-900 nanometers wavelengths, the bandpass varies 4-10 nanometers and involves 3-7 CCD rows.

The wedge-filter 110 includes wedge imaging spectrometer (WIS) technology, e.g., as developed by Santa Barbara Research Center (SBRC). See, G. Elerding, et al., "Wedge Imaging Spectrometer: Application to drug and pollution law enforcement", SPIE Vol. 1479, Surveillance Technologies (1991), pp. 380-392.

United States Patent 4,957,371, issued 9/18/90, to Pellicori, et al., also describes the implementation of a wedge-filter spectrometer.

The power conditioning unit 14, the temperature sensors 16, the cameras 18 and 20, the payload processor 22 and the solid- state memory 24 all formed a package housed in a lightweight aluminum box that weighed 8.07 kg in an ill-fated experimental system launched in October, 1995. The Conestoga launcher which carried the payload malfunctioned on its maiden voyage and was destroyed. The payload processor 22 comprises a Motorola R3081

RISC processor @ 25 MHz with R3000 CPU, R3010 FPU, twenty kilobyte cache, and sixteen megabytes of volatile RAM. The solid- state memory 24 provides forty megabytes of nonvolatile solid-state memory. The wedge filter camera 18 preferably provides both a 4.2° by 5.6° staring array monochrome imager and a visible /near-IR

"push-broom" hyperspectral system for 550-900 nanometers imaging. The VHR camera 20 preferably provides a field-of-view of 0.21° by 0.28°, 12.5/microradians per pixel, and includes the six- position optical-filter wheel 21. The instantaneous field-of-view (IFOV) of a pixel of the wedge filter camera 18 is preferably 255 microradians. From an altitude of 400 kilometers, each pixel of the wedge filter camera 18 is provided with a nominal nadir footprint of 102 meters. The single- image resolution is preferably limited not by the optics, but by sampling considerations, to the spacing of the pixels. The system has a positive contrast transfer function (CTF) for all spatial frequencies less than 510 microradians per line pair. Near the center of the field-of-view, the full-width at half maximum of the optical point-spread function is preferably about 1.5 pixels. The wedge filter camera 18 uses a silicon CCD focal plane array, e.g., a Thomson-CSF TH 7863 F02-01, equipped with a fiber bundle, unitary magnification, relay optic. Commercial manufacturers attach such fiber bundles directly to the CCD. The imaging surface is preferably the free, polished end of the bundle. The photo active area of the array has 288 rows and 384 columns of pixels on 23 micrometer square centers. A 4.2°x 5.6° field-of-view of the optics fills the CCD array. The CCD is preferably aligned with the columns (288 pixels each) in the cross-track direction. This yields a swath width of 29 kilometers cross-track and 39 kilometers in-track for nadir-viewing from an altitude of 400 kilometers.

The wedge filter, also called a linear variable filter, is preferably attached with transparent optical adhesive to the optical fiber bundle of the CCD. The filter has a nominal spectral resolution of 1.5% of the center wavelength, e.g., eight nanometers at 500 nanometers. The lines of constant wavelength, isochromes, are preferably parallel to the 288-pixel columns of the CCD, e.g., cross-track, and the spectral gradient is preferably parallel to the 384- pixel rows, e.g., in-track. Nominally, the wedge filter camera 18 spans the wavelength range from 550 to 900 nanometers, with the short-wavelength region being on the leading edge of the field-of- view and the long-wavelength region being on the trailing edge. The center of the local optical passband, the wavelength mean, varies approximately linearly along each row of the CCD at approximately one nanometer /pixel. The interval between successive images is preferably programmable in steps of 0.1 milliseconds. The maximum repetition rate is preferably limited by the processor protocol, e.g., one six frames per second, with twenty-five frames per second preferred. The wedge filter camera 18 has an analog-to-digital converter (ADC) with eight-bit resolution.

The wedge filter camera 18 has thirty-two commandable offsets, stepped from zero counts to nominally 248 counts, in increments of about eight counts. The wedge filter camera 18 has three programmable gain settings with relative gains that correspond approximately to 1050, 350, and 150 electrons per count. The integration time is preferably programmable between 0.4 and 773 millisecond in increments of 0.0944 milliseconds. The mean dark current per pixel generated from the camera at maximum gain and minimum offset is preferably less than one count/millisecond at focal plane temperatures less than 25°C. The dark current non- uniformity (one sigma) is preferably less than 10%.

For integration times less than twenty milliseconds and ambient temperatures of 2°-25°C, the frame-to-frame variation in dark count is preferably about 0.5 counts or less (RMS). This dark noise is the combined result of readout noise-current fluctuation and analog-to-digital round-off.

The instantaneous field-of-view (IFOV) of a single pixel of the VHR camera 20 is preferably 12.5 microradians square. From an altitude of 400 kilometers, each pixel of the VHR camera 20 has a nominal nadir footprint of five meters. The single-image resolution is not limited by the optics. Instead, the resolution is limited by the spacing of the pixels as a consequence of the sampling. The system has a positive contrast transfer function (CTF) for all spatial frequencies less than twenty-five microradians per line-pair. Near the center of the field-of-view, the full-width is about 1.1 pixels at half the maximum of the optical point-spread function.

The VHR camera 20 may use a silicon CCD array, e.g., a Thomson-CSF TH 7863 CRH-UV-01-B/T, with a proprietary coating to increase the sensitivity to the ultraviolet portions of the spectrum. The imaging surface is typically the silicon CCD itself. The photo-active area of the array is preferably 288 rows and 384 columns of pixels, e.g., on twenty-three micrometer square centers. Optics with a 0.21° by 0.28° field-of-view will fill the CCD array. The CCD is preferably aligned with the rows of 384 pixels each in the cross-track direction. This yields a swath width of 1.9 kilometers cross-track, and 1.4 kilometers in-track for nadir-viewing from 400 kilometers altitude.

The filter wheel 21 uses a six-position filter. The filters are arranged one every 60° of rotation, and the spectral bands of the filter wheel 21 are as shown in Table I, with full-width half- maximum (FWHM) nominal band values.

TABLE I

The filter wheel 21 step-and-settle time is preferably less than 250 milliseconds for steps between any two adjacent positions.

The mean-time-between-failure (MTBF) of the filter wheel 21 is preferably at least 3,000 hours of continuous operation.

The interval between successive images is preferably programmable in steps of one millisecond. The maximum repetition rate is preferably limited by the processor protocol, e.g., sixteen frames per second, with twenty-five frames per second preferred. An A/D converter typical included in the VHR camera 20 preferably has eight-bit resolution. The VHR camera 20 has thirty-two brightness offsets that step eight gray levels between adjacent offsets. The VHR camera 20 has three programmable gain settings with relative gains that correspond approximately to 1050, 350, and 150 electrons per count. The integration time is preferably programmable between 0.4 and 773 millisecond in increments of 0.0944 milliseconds. The mean dark current per pixel generated from the camera at maximum gain and minimum offset does not exceed one count/ms at focal plane temperatures less than 25° C. The dark current non-uniformity (one sigma) is preferably less than ten. At integration times less than 20 ms and ambient temperatures between 2°C and 25°C, the frame-to-frame variation in dark count is preferably about 0.5 counts or less (RMS). Such dark noise is preferably the combined result of readout noise, dark- current fluctuation, and analog-to-digital round-off.

The solid-state memory 24 includes 40M byte of non- volatile solid-state memory and thus can retain data when the power is off. The memory data transfer rate is at least 16K byte/second. In addition, 20M byte of volatile processor RAM is preferably available, 16M byte for image data and 4M byte for program. This memory cannot retain data when power is off. All image data passes through the RAM at a transfer rate of at least 2.5M bits per second.

The data rate from the sensor package to the low earth orbit experiment modulator 26 is preferably selectable at rates of 2.5M bits per second, 2.0M bits per second, or 1.66M bits per second. The interface pin out and connector type is preferred.

The total sensor package may be powered by a 28±4 volt direct current (DC) source. The twenty-eight volt return is preferably isolated from chassis ground. The power consumption of the package is preferably less than sixty watts. The worst case is expected to occur only when a filter wheel 21 is being turned and all US97/07814

-14-

the other electronic components are on. Maximum current is preferably 2.1 amps.

When the filter wheel 21 is not being repositioned, the power level should be less than 50 watts. An idle mode is preferably included that allows the cameras 18 and 20 and solid- state memory 24 to be turned completely off. The processor 22 and its RAM remain on. The power consumption in such idle mode is preferably less than thirty-five watts. The solid-state memory is powered from the same raw source that powers the filter wheel 21. Table II summarizes the power distribution.

TABLE II

The center of the field-of-view of the wedge filter camera 18 is preferably parallel to that of the very high resolution sensor within 0.8 milliradian (0.05°). Ground alignment tests show that a point that images at the center of the field-of-view of the very high resolution sensor, pixel (144,192), also images within three pixels of the center of the field-of-view of the wedge filter camera 18, pixel (144,192). The cameras 18 and 20 are mounted so their central lines of sight are perpendicular to the mounting plate of the package within two milliradians (0.12°).

The sensor package is preferably designed to operate over the temperature range of 10°C ± 3°C baseplate, provided that the mounting points on the spacecraft remain coplanar and in a fixed orientation with respect to the spacecraft's principal axes. The sensor package can operate with no harm to the electronics or any other part of the system over the temperature range of 2°C to 28°C.

The basic method of executing a sequence of commands involves sending script files. There are two types of commands: those common to all payloads and those specific to the payload processor 22. The payload processor 22 recognizes five basic commands from the satellite processor 32, as shown in Table III.

TABLE III

In addition to the basic commands, payload processor 22 preferably also recognizes a variety of payload-processor-specific commands which are interpreted and executed in script form by the processor 32. The use of canned scripts permits users to operate P 7 07814

-16-

either or both cameras 18 and 20 in an almost arbitrary interleaving of images with a wide variety of image rates.

The payload processor 22 preferably can receive revised software from the ground station, store it in the solid-state memory, and download it such that algorithms may be added, modified, or deleted in flight.

The satellite 12, to which a mounting plate for both the VHR and wedge-filter cameras 18 and 20 are attached, provides for aiming and ground communication with the VHR and wedge-filter cameras 18 and 20 from earth orbit. The satellite further provides for the orientation of the wedge-filter camera CCD 142 and wedge filter 140 relative to the direction of travel of the satellite platform in its earth orbit, e.g., orthogonal so that the CCD pixel rows are isochromes. Although particular embodiments of the present invention have been described and illustrated, such is preferably not intended to limit the invention. Modifications and changes will no doubt become apparent to those skilled in the art, and it is preferably intended that the invention only be limited by the scope of the appended claims.

Claims

THE INVENTION CLAIMED IS

1. A small-satellite imaging system, comprising: a very high resolution (VHR) camera fastened to a mounting plate such that its central line of sight is approximately perpendicular to the plane of said mounting plate, and comprising a beryllium telescope coupled to a silicon charge coupled device

(CCD) responsive to the visible and ultraviolet spectrums; a wedge-filter camera fastened to said mounting plate such that its central line of sight is also approximately perpendicular to the plane of said mounting plate, and comprising a beryllium telescope coupled to a silicon charge coupled device

(CCD) responsive to the far and near infrared spectrums with an interposed linear variable filter that passes light wavelengths to said CCD depending on the spatial orientation of the source to said central line of sight; and a satellite platform to which said mounting plate is attached and providing for aiming and ground communication with the VHR and wedge-filter cameras from earth orbit, and further providing for the orientation of said wedge-filter camera CCD relative to the direction of travel of the satellite platform in its earth orbit.

2. The system of Claim 1, wherein: the wedge filter camera has its center of the field-of-view parallel to that of the very high resolution camera within 0.8 milliradian (0.05°) such that a point that images at the center of the field-of-view of the very high resolution camera also images within three pixels of the center of the field-of-view of the wedge filter camera; and the wedge filter camera and the very high resolution camera are mounted such that their respective central lines of sight are perpendicular to the mounting plate within two milliradians

(0.12°).